JPMorgan’s AI Strategy: Chasing AI Dominance
摩根大通的人工智能战略：追求人工智能主导地位

The report explores JPMorgan’s AI strategy  to dominate the in AI.  The report includes the most exhaustive ai strategy analysis complete with references and works cited by Dany Kitishian of Klover.AI. 
该报告探讨了摩根大通在人工智能领域的战略，旨在占据主导地位。报告包含由 Klover.AI 的 Dany Kitishian 进行的最为详尽的人工智能战略分析，并附有参考文献和引用文献。

JPMorgan’s AI Strategy Executive Summary
摩根大通 AI 战略执行摘要

JPMorgan Chase (JPMC) is executing a comprehensive and aggressive artificial intelligence (AI) strategy that positions it not merely to compete in the ongoing technological revolution, but to achieve a decisive and lasting dominance in the financial services sector. This report provides an exhaustive analysis of JPMC’s approach, revealing that its preeminence is not a simple function of its massive financial outlay. Instead, it is the result of a cohesive, C-suite-driven strategy that masterfully integrates four mutually reinforcing pillars: a clear top-down mandate backed by unprecedented investment; a proprietary data and technology foundation built for scale; a broad portfolio of AI use cases delivering quantifiable returns across every business line; and a proactive, sophisticated approach to governance and talent that turns regulatory burdens and workforce transformation into competitive advantages.
摩根大通（JPMC）正在执行一项全面且积极的人工智能（AI）战略，该战略不仅使其能够参与正在进行的科技革命，更旨在实现金融服务业的决定性且持久的领先地位。本报告对摩根大通的策略进行了详尽分析，揭示其卓越地位并非仅仅是巨额资金投入的简单结果。相反，这是由一个协调一致、由高管团队主导的策略所驱动的，该策略巧妙地整合了四个相互强化的支柱：一个清晰的自上而下的指令，并得到前所未有的投资支持；一个为规模化而构建的专有数据和科技基础；一个涵盖各业务线的 AI 应用组合，能够带来可衡量的回报；以及一个积极主动、复杂的治理和人才策略，将监管负担和劳动力转型转化为竞争优势。

With an annual technology budget reaching $18 billion and a dedicated AI and data function reporting directly to CEO Jamie Dimon, the firm has elevated AI from a technology initiative to a core strategic imperative.¹ This commitment is projected to deliver $1.5 billion to $2.0 billion in tangible annual business value, a figure that continues to rise as the firm’s AI maturity deepens.³ The engine of this strategy is the firm’s “data flywheel,” a virtuous cycle powered by over $10 trillion in daily transactions, which provides an unparalleled source of high-stakes, proprietary data for training superior AI models.⁴ This data fuels a sophisticated ecosystem of in-house platforms like OmniAI and the LLM Suite, which democratize AI capabilities for over 200,000 employees and accelerate deployment at an industrial scale.⁵
该公司每年技术预算高达 180 亿美元，其专门的人工智能和数据部门直接向 CEO 杰米·戴蒙汇报，将人工智能从一项技术倡议提升为核心战略要务。 ¹ 这一承诺预计每年将为公司带来 15 亿至 20 亿美元的实际业务价值，随着公司人工智能成熟度的加深，这一数字还在持续增长。 ³ 该战略的核心是公司的“数据飞轮”，这是一个由超过 10 万亿美元每日交易驱动的良性循环，为公司提供了无与伦比的高风险、专有数据来源，用于训练更优秀的人工智能模型。 ⁴ 这些数据为 OmniAI 和 LLM 套件等内部平台组成的复杂生态系统提供动力，这些平台使超过 20 万名员工能够使用人工智能能力，并以工业规模加速部署。 ⁵

From automating 360,000 legal work hours annually with its COiN platform to boosting gross sales in Asset & Wealth Management by 20% with GenAI-driven tools, JPMC is systematically embedding AI into its operational fabric.⁷ This has resulted in significant, measurable returns, including holding fraud costs flat despite a 12% annual growth in attacks and increasing code deployments by over 70% in two years.⁵
从其 COiN 平台自动化每年 36 万小时的法务工作，到借助 GenAI 驱动工具将资产管理与财富管理的总销售额提升 20%，摩根大通正系统地将其运营体系中嵌入人工智能。 ⁷ 这带来了显著且可衡量的回报，包括在攻击量年增长 12%的情况下，将欺诈成本保持稳定，并在两年内将代码部署量增加超过 70%。 ⁵

Benchmarked against its Wall Street rivals, JPMC’s strategy stands apart in its scale and ambition. While competitors often focus on niche applications, JPMC is building an “AI factory” designed for enterprise-wide transformation, a fact validated by its consistent #1 ranking on the Evident AI Index for three consecutive years.⁹ The firm’s proactive investments in ethical AI, talent reskilling, and forward-looking research in quantum computing further cement its leadership position, creating a durable moat that will be difficult for competitors to breach. This report concludes that JPMC’s AI strategy is a masterclass in corporate transformation, providing a clear blueprint for how a legacy institution can leverage its scale to not only survive but dominate the AI era.
与华尔街竞争对手相比，摩根大通的战略在规模和雄心上独树一帜。虽然竞争对手通常专注于细分应用，但摩根大通正在建设一个“AI 工厂”，旨在实现企业范围内的转型，这一事实得到了其在连续三年稳居 Evident AI 指数榜首的验证。 ⁹ 该公司在伦理 AI、人才再培训和量子计算的前瞻性研究方面的积极投资，进一步巩固了其领导地位，构建了一个难以被竞争对手攻破的持久护城河。本报告总结认为，摩根大通的 AI 战略是企业转型的典范，为传统机构如何利用其规模不仅生存而且主导 AI 时代提供了清晰的蓝图。

JPMorgan Chase AI: Banking on AI Dominance

The Mandate for Dominance: Strategy, Investment, and Leadership
主导权使命：战略、投资与领导力

JPMorgan Chase’s foray into artificial intelligence is not a series of isolated experiments or a reactive measure to industry trends. It is a centrally orchestrated, top-down strategic imperative aimed at fundamentally re-architecting the firm for a new technological epoch. This mandate is defined by an unambiguous vision from the highest level of leadership, substantiated by a financial commitment of unprecedented scale, and executed through a governance structure built for speed and accountability.
摩根大通在人工智能领域的探索并非一系列孤立的实验或对行业趋势的被动应对。它是一项由高层主导、自上而下的战略强制要求，旨在从根本上重塑公司以适应新的技术时代。这一使命由最高领导层明确阐述的愿景所定义，通过前所未有的财务承诺得到支持，并通过一个为速度和问责制而构建的治理结构来执行。

The Dimon Doctrine: AI as a Transformational Force
戴蒙主义：AI 作为变革性力量

The driving force behind JPMC’s AI ambition is the clear and consistent vision articulated by its Chairman and CEO, Jamie Dimon. In his widely read annual letters to shareholders and public statements, Dimon has framed the impact of AI in historic terms, predicting it “will be extraordinary and possibly as transformational as some of the major technological inventions of the past several hundred years,” placing it alongside the printing press, electricity, and the internet.¹⁰ This is not mere corporate rhetoric; it is a strategic directive that permeates the firm’s culture and sets an urgent tone for organization-wide alignment and action.⁴
JPMC 追求 AI 雄心的驱动力，是其董事长兼 CEO 杰米·戴蒙所阐述的清晰且一贯的愿景。在他的广受关注的年度股东信和公开声明中，戴蒙将 AI 的影响置于历史背景下，预测它“将是非凡的，可能具有与过去几百年间一些重大技术发明同等变革性”，将其与印刷术、电力和互联网并列。 ¹⁰ 这不仅仅是企业空洞的言辞；它是一项战略指令，渗透到公司的文化中，为全组织的协调和行动设定了紧迫的基调。 ⁴

Dimon’s vision extends directly to the firm’s human capital, where he has been candid about the profound changes ahead. He has advised employees that the technology will not only augment their roles but may also “eliminate some careers”.⁴ This stark framing serves a crucial strategic purpose: it signals a non-negotiable shift in the firm’s operational fabric and preemptively manages expectations for the workforce transformation to come. By positioning AI as a critical determinant of future profitability and competitive advantage, the Dimon Doctrine ensures that the initiative receives the priority, resources, and cross-functional collaboration necessary to succeed.⁴
戴蒙的愿景直接延伸到公司的人力资本，他对即将到来的深刻变革直言不讳。他已建议员工，这项技术不仅会增强他们的角色，还可能“淘汰一些职业”。 ⁴ 这种鲜明的表述具有至关重要的战略意义：它标志着公司运营结构的不可协商转变，并预先管理了未来劳动力转型的预期。通过将人工智能定位为未来盈利能力和竞争优势的关键决定因素，戴蒙准则确保该计划获得必要的优先级、资源和跨职能协作以取得成功。 ⁴

The $18 Billion War Chest: Quantifying the Unprecedented Investment
180 亿美元的战争基金：量化前所未有的投资

JPMC substantiates its strategic vision with a financial commitment that dwarfs that of many competitors and even some dedicated technology firms. The bank’s total technology budget was approximately $17 billion in 2024, an amount that increased to a staggering $18 billion for 2025.¹ This level of spending, representing roughly 9.5% of the firm’s revenue, underscores the degree to which technology is viewed as a core business function rather than a support cost.⁹
JPMC 用一笔远超许多竞争对手甚至一些专门技术公司的资金投入来支撑其战略愿景。该银行 2024 年的总技术预算约为 170 亿美元，这一数字在 2025 年增加到了惊人的 180 亿美元。 ¹ 这种级别的支出，约占公司收入的 9.5%，突显了技术被视为核心业务功能而非支持成本的程度。 ⁹

A significant and growing portion of this budget is explicitly directed toward AI-centric initiatives. In 2024, approximately $1.3 billion was dedicated to advancing AI capabilities.⁹ This focus is evident across business lines; the Consumer & Community Banking (CCB) division alone plans to spend approximately $9 billion on technology, product, and design in 2025.⁴ This spending is not speculative but is tied to clear financial objectives. Daniel Pinto, the firm’s President and Chief Operating Officer, projects that AI use cases will deliver between $1.5 billion and $2.0 billion in tangible business value annually. This projection represents an increase from an earlier estimate of $1.0 billion to $1.5 billion, signaling growing confidence as initiatives mature and deliver results.³ Fraud prevention has been cited as a particularly significant contributor to this value creation.³
这部分预算中一个重要且不断增长的部分明确用于以 AI 为中心的举措。2024 年，大约有 13 亿美元专门用于提升 AI 能力。 ⁹ 这种关注体现在各个业务线；仅消费与社区银行（CCB）部门就计划在 2025 年投入约 90 亿美元用于技术、产品和设计。 ⁴ 这项支出并非投机行为，而是与明确的财务目标紧密相连。公司总裁兼首席运营官 Daniel Pinto 预计，AI 应用案例每年将带来 15 亿至 20 亿美元的实际业务价值。这一预测是从早先的 10 亿至 15 亿美元估计中增加的，显示出随着举措成熟并取得成果，信心日益增强。 ³ 防欺诈已被列为这一价值创造的特别重要贡献者。 ³

This aggressive investment strategy has fundamentally shifted from defense to offense. CFO Jeremy Barnum has indicated that the firm’s “modernization investment has peaked,” meaning the most difficult and costly foundational work of retiring technical debt and migrating to the cloud is well underway.⁴ Unlike competitors still mired in these earlier stages, JPMC is now able to reallocate its massive budget from defensive overhauls to offensive, innovation-driven AI initiatives. The firm is now “harvesting the fruits of that labor to fund its quest for AI dominance,” creating an accelerating, compounding lead over rivals who are still paying to build the foundational “highway” while JPMC deploys a fleet of high-performance AI “race cars” onto its already-built infrastructure.⁴
这种激进的投入策略已经从根本上从防御转向进攻。首席财务官杰里米·巴 num 表示，公司“现代化投资已达到顶峰”，这意味着最困难、最昂贵的退休技术债务和迁移到云的基础性工作已经顺利展开。 ⁴ 与仍深陷于这些早期阶段的竞争对手不同，摩根大通现在能够将其庞大的预算从防御性改造重新分配到由创新驱动的进攻性人工智能计划上。该公司现在“正在收获这些劳动的成果，以资助其在人工智能领域的统治地位追求”，在那些仍在支付费用以建设基础“高速公路”的竞争对手中，摩根大通已经将其高性能人工智能“赛车队”部署到了已经建成的基础设施上，从而创造了加速、复利的领先优势。 ⁴

The AI Power Structure: Governance Built for Execution
人工智能权力结构：为执行而构建的治理

To steer this monumental effort, JPMC has institutionalized its AI strategy through a dedicated and powerful leadership structure designed to overcome the bureaucratic inertia that often plagues large organizations. A landmark decision was the 2023 appointment of Teresa Heitsenrether as the firm’s first Chief Data & Analytics Officer.³ Crucially, Heitsenrether sits on the firm’s Operating Committee, granting her the executive authority and visibility to drive the AI agenda across all lines of business, establish firm-wide data governance standards, and spearhead AI adoption to improve productivity and risk management.⁴
为了领导这项宏伟计划，摩根大通通过一个专门且强大的领导结构来制度化其人工智能战略，旨在克服大型组织常受的官僚主义惰性。一个里程碑式的决策是在 2023 年任命特蕾莎·海茨纳雷瑟为公司的首位首席数据与分析官。 ³ 关键在于，海茨纳雷瑟坐在公司的运营委员会上，赋予她推动跨所有业务领域的人工智能议程、建立全公司数据治理标准以及主导人工智能应用以提高生产力和风险管理的高层执行权和可见度。 ⁴

In what can be seen as a deliberate redesign of the corporate operating system for the AI age, the firm undertook a critical organizational restructuring. As Jamie Dimon confirmed at the Databricks Data + AI Summit, “We took AI, slash data, out of technology. It’s too important”.² The AI and data function was moved from the traditional technology hierarchy to report directly to Dimon and the company president.² This structural change is a competitive weapon. By elevating AI to the highest level of management attention, JPMC ensures it receives priority, firm-wide alignment, and the necessary resources to succeed. It breaks down the silos that typically cause technology initiatives to become bogged down within the IT department, reframing AI from a “cost center” to a “value generator” and dramatically accelerating decision-making and deployment.
可以看作是为适应人工智能时代而对企业操作系统进行有意重新设计的举措，该公司进行了一次关键的组织结构调整。正如杰米·戴蒙在 Databricks 数据+人工智能峰会上的确认，“我们将人工智能、数据从技术中分离出来。这太重要了”。 ² 人工智能和数据职能从传统技术层级体系中分离出来，直接向戴蒙和公司总裁汇报。 ² 这一结构性变化是一种竞争武器。通过将人工智能提升到管理层最高级别的关注，摩根大通确保其获得优先权、全公司的协同一致以及成功所需的必要资源。它打破了通常导致技术计划在 IT 部门陷入困境的壁垒，将人工智能从“成本中心”重新定义为“价值生成器”，并极大地加速了决策和部署。

This C-suite leadership is supported by a deep bench of world-class talent. The J.P. Morgan AI Research Lab is headed by Dr. Manuela Veloso, a globally renowned AI expert formerly of Carnegie Mellon University, who leads a team of over 200 researchers exploring the frontiers of the field.⁴ Complementing this research focus is a pragmatic approach to policy and compliance, led by Terah Lyons, the global head of AI and Data Policy, whose role includes engaging directly with regulators to shape and understand the evolving compliance landscape.⁴ This combination of executive power, research prowess, and policy expertise creates a governance structure that is built for rapid, responsible, and scalable execution.
这位高管团队的领导力得到了世界级人才库的坚实支持。摩根大通人工智能研究实验室由卡内基梅隆大学出身的全球知名人工智能专家、曼努埃拉·维洛索博士领导，她带领一个由 200 多名研究人员组成的团队，探索该领域的最前沿。 ⁴ 与这一研究重点相补充的是，由人工智能与数据政策全球负责人泰拉·莱昂斯领导的政策与合规方面的务实方法，她的职责包括直接与监管机构互动，塑造并理解不断发展的合规环境。 ⁴ 这种高管权力、研究实力和政策专业知识的结合，创造了一个为快速、负责任和可扩展执行而构建的治理结构。

JPMC AI Investment & Leadership Scorecard
摩根大通人工智能投资与领导力评分卡

Metric  指标	Value / Target  价值/目标	Key Executive(s) Responsible 负责高管	Source(s)  来源
Annual Technology Budget (2025) 年度技术预算（2025）	~$18 billion  ~180 亿美元	Jeremy Barnum (CFO)  杰里米·巴 num（首席财务官）	1
Projected Annual AI-Related Business Value 预计年人工智能相关业务价值	$1.5 billion – $2.0 billion 1.5 亿至 2 亿美元	Daniel Pinto (President & COO) 丹尼尔·平托（总裁兼首席运营官）	3
CCB Tech, Product & Design Spend (2025) CCB 技术、产品与设计支出（2025 年）	~$9 billion  ~90 亿美元	Marianne Lake (CEO, CCB)  玛丽安·莱克（CCB 首席执行官）	4
AI & Machine Learning Specialists (Current) AI 与机器学习专家（现任）	2,000+	Teresa Heitsenrether (CDAO) 特蕾莎·海茨纳雷瑟（CDAO）	10
AI & Machine Learning Specialists (Target) AI & 机器学习专家（目标）	5,000	Jamie Dimon (CEO)  杰米·戴蒙（CEO）	15
AI Researchers  AI 研究人员	200+	Dr. Manuela Veloso (Head of AI Research) 曼努埃拉·维洛索博士（AI 研究主管）	4

The Engine Room: Building an “AI-Ready” Foundation
引擎室：构建“AI 就绪”的基础设施

JPMorgan Chase’s AI strategy is not built on algorithms alone. It rests upon a meticulously engineered and modernized technical foundation that serves as the firm’s most defensible competitive moat. This engine room consists of three critical components: an unparalleled proprietary data asset, a modernized cloud and data infrastructure designed to make that data usable, and a suite of in-house platforms that democratize and scale AI development across the enterprise.
摩根大通的人工智能战略并非仅基于算法。它建立在经过精心设计和现代化的技术基础设施之上，构成了公司最难以逾越的竞争壁垒。这个引擎室由三个关键组成部分构成：无与伦比的专有数据资产、现代化的云和数据基础设施，旨在使这些数据变得可用，以及一系列内部平台，这些平台在企业范围内民主化和扩展人工智能开发。

The Data Flywheel: The Unparalleled Advantage of Proprietary Data
数据飞轮：专有数据的无与伦比优势

At the heart of JPMC’s competitive strategy lies its most profound advantage: a virtuous, self-reinforcing “data flywheel”.⁴ This flywheel is fueled by a daily transactional flow of staggering proportions, exceeding $10 trillion across more than 120 currencies and 160 countries.⁴ This is not the passive, low-value data of clicks and social media likes that fuels many tech companies. It is high-velocity, “high-stakes” transactional data that provides a rich, real-time, and granular view of the global economy.⁴
JPMC 的竞争策略的核心在于其最深刻的优势：一个良性循环、自我强化的“数据飞轮”。 ⁴ 这个飞轮由每天高达惊人的交易流量驱动，跨越 120 多种货币和 160 多个国家，交易总额超过 10 万亿美元。 ⁴ 这不是许多科技公司依赖的被动、低价值的点击和社交媒体点赞数据。它是一种高速度、高风险的交易数据，提供了对全球经济丰富、实时和细粒度的洞察。 ⁴

This asset, which Marianne Lake, CEO of Consumer & Community Banking, aptly describes as a “very rich and valuable tapestry of data,” encompasses the full spectrum of financial activity.⁴ It includes 34 million customer incomes, 36 million credit profiles, 18 billion annual digital log-ins, and 25 billion annual credit and debit card transactions.¹⁷ The power of this data advantage stems not just from its sheer volume but from its inherent consequence. Every data point processed by JPMC is tied to a real, regulated, and often high-value financial transaction. The feedback loop generated by this data is therefore incredibly powerful and direct. When an AI model is trained on this data for a task like credit risk assessment or fraud detection, its performance has immediate and substantial real-world consequences, forcing a degree of rigor and creating models that are battle-tested in a way that those trained on lower-stakes data are not.⁴ These superior models, in turn, enhance products and services, leading to better customer outcomes, which attracts more business and generates an even greater volume of high-quality, proprietary data, thus powering the flywheel.⁴
这项资产，正如消费者与社区银行首席执行官 Marianne Lake 恰如其分地描述的那样，是一个“非常丰富和有价值的数据织锦”，涵盖了所有金融活动的全范围。 ⁴ 它包括 3400 万客户收入、3600 万信用档案、180 亿年数字登录次数以及 250 亿年信用卡和借记卡交易。 ¹⁷ 这种数据优势的力量不仅源于其庞大的数量，更源于其内在的后果。JPMC 处理的每一个数据点都与真实的、受监管的、通常是高价值的金融交易相关联。因此，这种数据产生的反馈循环非常强大和直接。当 AI 模型基于这些数据训练用于信用风险评估或欺诈检测等任务时，其表现会立即产生实质性的现实后果，从而迫使一种严谨性，并创造出那些在低风险数据上训练的模型所不具备的实战检验模型。 ⁴ 这些更优秀的模型反过来又提升了产品和服务，带来了更好的客户结果，从而吸引更多业务并产生更大量的高质量专有数据，从而驱动飞轮。 ⁴

 From Legacy to AI-Native: Infrastructure Modernization
从传统到 AI 原生：基础设施现代化

To harness the power of its data flywheel, JPMC has embarked on a multi-year effort to modernize its vast technology estate and make its data “AI Ready”.¹ A critical focus of this initiative is to deliver and curate data assets so they are discoverable, highly accurate, and available in milliseconds, pulling humans out of the laborious cycle of fixing and preparing data.¹ This modernization has been aggressive, with the firm reporting that approximately 65% of its applications are now running a large part of their workloads on the public or private cloud, a significant increase from 50% in the preceding year.¹
为了发挥其数据飞轮的强大力量，摩根大通正致力于进行为期多年的现代化技术资产改造，使其数据达到“AI 就绪”状态。 ¹ 该计划的重点之一是提供和整理数据资产，确保它们可发现、高度准确，并且能在毫秒级内获取，从而将人类从繁琐的数据修复和准备工作中解放出来。 ¹ 这次现代化改造非常激进，公司报告称，目前约 65%的应用程序正在公共云或私有云上运行其大部分工作负载，与上一年 50%的比例相比有了显著提升。 ¹

The cornerstone of the firm’s data initiatives is the JPMorgan Chase Advanced Data Ecosystem (JADE) platform. JADE serves as the foundation for AI initiatives by providing the high-quality, unified data that is essential for training and deploying sophisticated models.⁹ To accelerate AI development at scale, JPMC has strategically implemented a data mesh architecture. This approach decentralizes data into product-specific data lakes and utilizes tools like the AWS Glue Data Catalog, allowing teams across different business units to quickly access and experiment with relevant datasets for model training.⁹ This distributed architecture has proven crucial for developing specialized AI solutions, such as fraud detection models requiring real-time access to transaction data, while maintaining robust data governance and security.⁹
该公司数据计划的核心是摩根大通高级数据生态系统（JADE）平台。JADE 通过提供用于训练和部署复杂模型所必需的高质量、统一数据，为人工智能计划奠定基础。 ⁹ 为了大规模加速人工智能开发，摩根大通战略性地实施了数据网格架构。这种方法将数据分散到特定产品的数据湖中，并利用 AWS Glue 数据目录等工具，使不同业务部门的团队能够快速访问和实验相关数据集以进行模型训练。 ⁹ 这种分布式架构对于开发专业人工智能解决方案至关重要，例如需要实时访问交易数据的欺诈检测模型，同时保持强大的数据治理和安全。 ⁹

The Platform Play: Democratizing and Scaling AI
平台战略：民主化与扩展 AI

JPMC’s strategy recognizes that even the best data and infrastructure are insufficient without platforms to scale deployment and adoption. The firm has developed a suite of proprietary platforms that tackle the biggest barriers to enterprise AI. This dual-platform strategy addresses both the technology stack and the human stack simultaneously, dramatically increasing the probability of successful, at-scale deployment.
摩根大通的战略认识到，没有平台来扩展部署和采用，即使最好的数据和基础设施也是不足的。该公司开发了一套专有平台，以解决企业 AI 面临的最大障碍。这种双平台战略同时解决技术栈和人类栈问题，极大地提高了大规模成功部署的可能性。

First, OmniAI is a major in-house innovation that serves as a critical integration layer for the entire firm.⁶ Developed in the Chief Technology Office, OmniAI solves the immense technical challenge of applying modern AI and machine learning models on top of the firm’s complex legacy infrastructure. It standardizes processes and provides the necessary security and controls for working in a highly regulated environment, enabling developers to work faster and deliver more AI-powered applications to the firm’s businesses.⁶
首先，OmniAI 是一项重要的内部创新，作为整个公司的关键集成层。 ⁶ OmniAI 由首席技术官办公室开发，解决了在现代 AI 和机器学习模型上应用公司复杂遗留基础设施的巨大技术挑战。它标准化了流程，并为在高度监管的环境中工作提供了必要的安全和控制，使开发人员能够更快地工作，并向公司的业务交付更多 AI 驱动的应用程序。 ⁶

Second, the LLM Suite is a model-agnostic generative AI platform that solves the human problem of adoption and upskilling. It has been rolled out to over 200,000 employees globally, providing them with a secure, in-house alternative to public tools like ChatGPT.¹ The platform acts as a “virtual research analyst,” enabling employees to summarize documents, generate ideas, and query knowledge bases. This is already saving many users “several hours per week” on less valuable tasks and is fostering a culture of innovation by enabling “citizen developer” use cases to go into production.⁵
其次，LLM 套件是一个模型无关的生成式 AI 平台，它解决了人类在采用和提升技能方面的问题。该套件已在全球范围内推广给超过 20 万名员工，为他们提供了对 ChatGPT 等公共工具的安全、内部替代方案。 ¹ 该平台充当“虚拟研究分析师”，使员工能够总结文档、生成想法和查询知识库。这已经帮助许多用户在价值较低的任务上节省了“每周数小时”的时间，并通过支持“公民开发者”用例进入生产，培养了创新文化。 ⁵

This platform strategy is augmented by strategic partnerships with key technology providers. The firm is a customer of Databricks for advanced analytics and machine learning operations, leverages AWS for cloud infrastructure, and maintains close partnerships with hardware manufacturers to secure its supply of compute capacity and “control our own destiny” in a resource-constrained environment.⁹
这一平台战略通过与技术提供商的战略合作得到增强。该公司的先进分析和机器学习运营业务使用 Databricks，利用 AWS 进行云基础设施，并与硬件制造商保持密切合作，以确保其计算能力的供应，并在资源受限的环境中“掌控自己的命运”。 ⁹

JPMC Proprietary AI Platform Ecosystem
JPMC 专有 AI 平台生态系统

Platform Name  平台名称	Core Function  核心功能	Primary Business Line(s)  主要业务线	Key Metric / User Base 关键指标/用户基础	Source(s)  来源
JADE	AI-Ready Data Foundation  AI 准备数据基础	Firm-wide (Technology)  全公司（技术）	Cornerstone of AI initiatives, provides unified data AI 计划的核心，提供统一数据	9
OmniAI  全维人工智能	AI/ML Integration & Deployment AI/ML 集成与部署	Firm-wide (Technology)  全公司范围（技术）	Accelerates AI deployment on legacy systems 加速在传统系统上的 AI 部署	4
LLM Suite  LLM 套件	Employee Productivity & GenAI Access 员工生产率与生成式 AI 访问	Firm-wide  全公司范围	200,000+ employee users  20 万+员工用户	1
COiN	Legal Document Automation 法律文件自动化	Legal, Compliance, Operations 法律、合规、运营	Saves 360,000+ legal hours annually 每年节省超过 36 万小时的法律工作	7
IndexGPT	Thematic Investment Product Creation 主题投资产品创建	Asset & Wealth Management 资产管理与财富管理	Uses GPT-4 for new index generation 使用 GPT-4 进行新指数生成	21
EVEE Intelligent Q&A  EVEE 智能问答	Call Center Agent Support 客服中心代理支持	Consumer & Community Banking 消费者与社区银行	Improves call resolution and agent efficiency 提高通话解决率和代理效率	23
J.P. Morgan Virtual Assistant 摩根大通虚拟助手	Client Self-Service (J.P. Morgan Access) 客户自助服务（摩根大通接入）	Corporate & Investment Bank 企业及投资银行	Real-time transaction queries and wire tracking 实时交易查询与电汇追踪	24
“Ask David”  “问大卫”	Investment Research Automation 投资研究自动化	Private Bank  私人银行	Multi-agent system with >90% query accuracy 具有>90%查询准确率的分布式代理系统	4

AI in Action: A Cross-Enterprise Portfolio of Use Cases
AI 应用：跨企业用例组合

JPMorgan Chase is systematically embedding artificial intelligence across every major business line, moving far beyond experimentation to create a comprehensive portfolio of applications that are in production and delivering tangible value. The firm’s strategy demonstrates both breadth—touching all segments from corporate functions to consumer banking—and depth, with AI being applied to enhance efficiency, mitigate risk, and generate revenue. The firm currently has over 600 AI use cases in production, a number that has grown rapidly and reflects a mature, enterprise-wide implementation.²
摩根大通正在系统地将其人工智能嵌入到每一个主要业务领域，远超实验阶段，创建了一组全面的应用程序，这些应用程序已投入生产并产生实际价值。该公司的战略展示了其广度——涵盖从企业职能到消费者银行的所有领域——和深度，人工智能被应用于提高效率、降低风险和创造收入。该公司目前已有超过 600 个 AI 应用案例在生产中，这一数字增长迅速，反映了成熟的企业级实施。 ²

Corporate & Shared Services: Automating the Core
企业共享服务：核心业务自动化

The firm’s AI strategy begins with its own internal operations, where automation and intelligence are being applied to create a more efficient and effective corporate core. This “inside-out” approach is strategically sound, as it allows JPMC to master the technology, governance, and workflows in a controlled environment. The cost savings and productivity gains from these internal applications effectively de-risk and subsidize the R&D for future client-facing products, creating a sustainable innovation engine.
该公司的 AI 战略始于其内部运营，通过自动化和智能化应用，打造更高效、更有效的企业核心。这种“由内而外”的方法具有战略优势，因为它允许摩根大通在受控环境中掌握技术、治理和业务流程。这些内部应用带来的成本节约和生产力提升，有效降低了面向客户产品的研发风险并提供了资金支持，构建了一个可持续的创新引擎。

• Legal & Compliance (COiN): Perhaps the most celebrated example of JPMC’s operational AI is the Contract Intelligence (COiN) platform. This system leverages Natural Language Processing (NLP) and machine learning to analyze and extract key data points from complex legal documents. COiN is capable of processing 12,000 commercial credit agreements in a matter of seconds, a task that previously required immense manual effort. The platform has delivered dramatic results, saving over 360,000 hours of legal and loan officer work annually.⁷ Beyond time savings, it has significantly improved accuracy, reducing compliance-related errors by an estimated
  法律与合规（COiN）：JPMC 运营 AI 最著名的例子可能是合同智能（COiN）平台。该系统利用自然语言处理（NLP）和机器学习来分析和提取复杂法律文件中的关键数据点。COiN 能够在几秒钟内处理 12,000 份商业信贷协议，而这项任务以前需要大量人工工作。该平台已取得显著成果，每年节省超过 360,000 小时的法律和贷款员工作时间。 ⁷ 除了节省时间，它还显著提高了准确性，据估计减少了与合规相关的错误。
  80% and lowering the overall cost of these legal operations by 30%.⁷
  提升效率 80%并降低这些法律操作的整体成本 30%。 ⁷
• Software Development: JPMC is applying AI to its own technology development lifecycle. Over 40,000 engineers now have access to AI coding assistants.¹⁴ These tools are boosting developer efficiency by
  软件开发：摩根大通正将其 AI 技术应用于自身的技术开发生命周期。目前超过 40,000 名工程师可以访问 AI 编码助手。 ¹⁴ 这些工具正在通过
  10-20%, automating tasks like code debugging, optimization, and even the complex migration of legacy systems from COBOL to Java.²³ This has been a key driver behind a
  10-20%，自动化代码调试、优化，甚至从 COBOL 到 Java 的遗留系统复杂迁移等任务。 ²³ 这一直是其背后的关键驱动因素。
  70% increase in code deployments over the last two years and a 20% reduction in work being replanned, accelerating the delivery of new products and features across the bank.⁵
  过去两年代码部署量增长了 70%，重新规划的工作量减少了 20%，加速了银行新产品和功能的交付。 ⁵
• Human Resources: The firm is also deploying AI in its HR functions to improve fairness and efficiency. One patented system provides for the automated anonymization of resumes, redacting identifiers like names and universities to force hiring managers to focus solely on skills and mitigate unconscious bias.²⁶
  人力资源：该机构也在其人力资源职能中部署人工智能，以提高公平性和效率。一个已获专利的系统提供简历的自动匿名化，删除姓名和大学等标识符，迫使招聘经理仅关注技能并减轻无意识偏见。 ²⁶

Consumer & Community Banking (CCB): Fortifying the Front Lines
消费者与社区银行（CCB）：巩固前线

In its largest division, which serves 84 million US customers, JPMC is using AI to enhance security, improve customer service, and deliver more personalized experiences.⁵
在其规模最大的部门，该部门服务于 8400 万美国客户，摩根大通正利用人工智能来增强安全性、改善客户服务，并提供更加个性化的体验。 ⁵

• Fraud Detection & Risk Management: AI is the cornerstone of the bank’s defense against financial crime. The firm faces a 12% compound annual growth rate in fraud attacks, yet its AI-powered tools have enabled it to hold the cost of fraud flat—a remarkable achievement that represents a massive cost avoidance.⁵ Machine learning models analyze millions of transactions in real-time with 98% accuracy, evaluating behavioral signals like typing cadence to detect credit card fraud and using NLP to identify business email compromise scams.²³ In anti-money laundering (AML) surveillance, AI has achieved a
  欺诈检测与风险管理：AI 是银行防范金融犯罪的核心。该机构面临欺诈攻击复合年增长率达 12%的挑战，但其 AI 驱动的工具使其能够将欺诈成本保持稳定——这是一个显著的成就，代表着巨大的成本规避。 ⁵ 机器学习模型实时分析数百万笔交易，准确率达 98%，通过评估打字节奏等行为信号来检测信用卡欺诈，并利用自然语言处理技术识别商业电子邮件妥协诈骗。 ²³ 在反洗钱（AML）监控中，AI 已取得
  95% reduction in false positives, allowing investigators to focus on genuine threats.⁹
  将错误警报减少 95%，使调查人员能够专注于真实威胁。 ⁹
• Call Center & Operations Automation: To improve the efficiency of its vast customer service operations, JPMC has deployed the EVEE Intelligent Q&A generative AI assistant. This tool provides call center agents with instant, context-aware responses to customer inquiries by integrating with policy documents and transaction histories, leading to faster resolution of issues and higher customer satisfaction.²³ This, combined with broader process automation, has contributed to keeping non-interest expenses flat in the CCB division over the last five years, even as business volumes have grown. Key metrics demonstrate this impact: servicing calls per account are down nearly 30%, and processing costs are down 15%.⁵
  呼叫中心与运营自动化：为提升其庞大的客户服务运营效率，摩根大通部署了 EVEE 智能问答生成式 AI 助手。该工具通过与政策文件和交易历史集成，为呼叫中心代理提供即时、上下文感知的回应，从而加快问题解决速度并提高客户满意度。 ²³ 结合更广泛的过程自动化，这有助于在业务量增长的同时，使 CCB 部门的非利息支出在过去五年中保持稳定。关键指标显示了这一影响：每账户服务电话量下降了近 30%，处理成本降低了 15%。 ⁵
• Customer Personalization & Experience: The firm is leveraging its vast data on shopping behavior and digital engagement to deliver personalized products and services.¹⁷ For corporate clients using the J.P. Morgan Access® platform, the
  客户个性化与体验：该机构正利用其庞大的购物行为和数字互动数据来提供个性化产品和服务。 ¹⁷ 对于使用 J.P. Morgan Access®平台的 corporative 客户，
  J.P. Morgan Virtual Assistant uses AI to provide real-time answers to queries like “What wires cleared yesterday?” and can track international payments across correspondent banks, eliminating the need for manual follow-up.²⁴
  摩根大通虚拟助手运用人工智能技术，可实时回答“昨日哪些电汇已清算？”等查询，并能追踪通过代理银行进行的国际支付，无需人工跟进。 ²⁴

The Corporate & Investment Bank (CIB): Enhancing High-Value Decisions
企业及投资银行（CIB）：提升高价值决策

Within the CIB, AI is being deployed to augment the capabilities of bankers and traders, automating research and enhancing decision-making in high-stakes environments.
在企业及投资银行部门，人工智能被用于增强银行家和交易员的能力，自动化研究并在高风险环境中提升决策水平。

• Algorithmic Trading: JPMC has been using AI on its trading desks for over eight years.⁵ AI-powered trading algorithms analyze market data, identify opportunities, and execute trades at optimal times.²⁷ The firm has used reinforcement learning to improve trading win rates from 52% to 63% and save an estimated $25 million in slippage costs through optimized order routing.²⁷ Today, over 50% of the bank’s electronic Foreign Exchange (eFX) Spot volume in sizes greater than $10 million is handled by its proprietary FX Algos.⁴
  算法交易：摩根大通的交易部门已使用人工智能超过八年。 ⁵ 人工智能驱动的交易算法分析市场数据，识别机会，并在最佳时机执行交易。 ²⁷ 该公司通过强化学习将交易胜率从 52%提升至 63%，并通过优化的订单路由节省了约 2500 万美元的滑点成本。 ²⁷ 目前，超过 50%的摩根大通电子外汇（eFX）现货交易量（交易金额超过 1000 万美元）由其专有的外汇算法处理。 ⁴
• Banker Productivity & Client Prospecting: AI tools are dramatically increasing the efficiency of investment bankers. These tools can automate up to 40% of research tasks by summarizing SEC filings, digesting millions of call reports, and generating initial valuation models.¹ This frees up bankers to focus on strategic advice and client relationships. AI is also being used for growth, with models analyzing data to identify a prospect list of
  银行生产力与客户开发：AI 工具正在极大地提高投资银行的工作效率。这些工具能够通过总结 SEC 文件、消化数百万份电话报告以及生成初步估值模型来自动化高达 40%的研究任务。 ¹ 这让银行家能够专注于战略建议和客户关系。AI 也被用于增长，通过分析数据来识别潜在客户名单。
  60,000 middle-market companies and to flag existing clients who may be ready to do a bond offering, for example.¹
  60,000 家中型市场公司，并标记可能已准备好进行债券发行的现有客户，例如。 ¹
• Risk Management: The Morpheus system is an AI-powered platform that analyzes massive datasets to identify and manage risks associated with complex trades and financial models, enhancing the firm’s financial stability.²⁰
  风险管理：Morpheus 系统是一个 AI 驱动的平台，通过分析海量数据来识别和管理复杂交易与金融模型相关的风险，从而增强公司的财务稳定性。 ²⁰

Asset & Wealth Management (AWM): Empowering Advisors
资产管理与财富管理（AWM）：赋能顾问

In the AWM division, AI is being used to empower financial advisors with better tools, leading to more personalized client advice and the creation of innovative investment products. This creates a reinforcing loop across business lines, where the anonymized, aggregated data on consumer spending trends from CCB can provide unique, proprietary insights into emerging themes. This data can then inform the “themes” that AWM’s IndexGPT identifies and builds products around, creating a cross-silo advantage that is difficult for competitors to replicate.
在 AWM 部门，人工智能正被用来为财务顾问提供更好的工具，从而实现更个性化的客户建议和创新投资产品的开发。这形成了一个跨业务线的强化循环，其中来自 CCB 的消费支出趋势的匿名化、聚合数据能够提供独特的、专有的洞察力，揭示新兴主题。这些数据随后将指导 AWM 的 IndexGPT 识别和围绕这些“主题”构建产品，从而创造一个难以被竞争对手复制的跨部门优势。

• Advisor Productivity Tools: The Coach AI tool acts as a real-time assistant for wealth managers, enabling them to instantly access research, market trends, and personalized investment recommendations. During periods of market volatility, this tool has proven invaluable, helping advisors find the right information for client conversations up to 95% faster.⁸ This efficiency gain is expected to allow advisors to significantly scale their business, with projections that the tool will help them expand their client rosters by
  顾问生产力工具：教练 AI 工具作为财富管理人员的实时助手，使他们能够即时获取研究、市场趋势和个性化投资建议。在市场波动期间，该工具已被证明极具价值，帮助顾问在客户对话中找到正确信息，速度提高了 95%。 ⁸ 这种效率提升预计将使顾问显著扩大其业务规模，预计该工具将帮助他们扩大客户群体。
  50% in the next three to five years.⁸
  未来三到五年内达到 50%。 ⁸
• Personalized Investing & New Products: JPMC has demonstrated its ability to innovate with the launch of IndexGPT, a tool that uses OpenAI’s GPT-4 model to create thematic investment baskets.¹⁵ This application moves beyond simple automation to generate novel investment products tailored to emerging trends.
  个性化投资与新产品：摩根大通通过推出 IndexGPT 展现了其创新能力，该工具利用 OpenAI 的 GPT-4 模型创建主题投资组合。 ¹⁵ 这项应用超越了简单的自动化，能够生成针对新兴趋势的定制化投资产品。
• Advanced Research Automation: Pushing the boundaries of automation, the Private Bank has developed “Ask David,” a sophisticated multi-agent AI system designed to automate complex investment research tasks. The system has demonstrated over 90% accuracy on domain-specific queries, showcasing the firm’s ability to build highly specialized and effective AI solutions.⁴
  高级研究自动化：私人银行在自动化领域拓展了边界，开发了“Ask David”——一个复杂的、基于多智能体的 AI 系统，旨在自动化复杂的投资研究任务。该系统在特定领域查询中展现出超过 90%的准确率，展示了该机构构建高度专业化和有效的 AI 解决方案的能力。 ⁴

The Bottom Line: Quantifying the Return on AI Investment
核心要点：量化人工智能投资的回报

JPMorgan Chase’s multi-billion-dollar investment in artificial intelligence is not an act of faith; it is a calculated strategy designed to produce substantial and measurable returns. The firm has been notably transparent in quantifying the impact of its AI initiatives, signaling a high degree of confidence to investors and setting a benchmark for the industry. The returns manifest across three key areas: direct financial gains and ROI, transformative efficiency and productivity improvements, and tangible contributions to revenue growth.
摩根大通数十亿美元的人工智能投资并非出于信念；而是一项精心策划的策略，旨在产生巨大且可衡量的回报。该公司在量化其人工智能计划的影响方面表现出显著的透明度，向投资者传递出高度自信的信号，并为行业树立了标杆。回报体现在三个关键领域：直接的财务收益和投资回报率、变革性的效率和生产力提升，以及对公司收入增长的实质性贡献。

The ROI Scorecard: Hard Financial Metrics
投资回报率评分卡：硬性财务指标

Unlike many peers who speak of AI in abstract terms, JPMC has consistently provided concrete financial metrics to demonstrate the value of its technology spend. The firm is one of the few major financial institutions to publicly report on the return on investment (ROI) from its AI use cases, a move that underscores the maturity of its program.²⁹
与许多只停留在抽象谈论 AI 的同行不同，摩根大通一直通过具体的财务指标来展示其技术投入的价值。该机构是少数几家公开报告其 AI 应用案例投资回报率（ROI）的主要金融机构之一，这一举措凸显了其项目的成熟度。

The headline figure, projected by President and COO Daniel Pinto, is that AI initiatives are on track to deliver $1.5 billion to $2.0 billion in annual business value.³ This value is derived from a combination of revenue uplift, expense reduction, and cost avoidance. The firm’s confidence in this return is growing; this projection is an increase from a previous estimate of $1.0 billion to $1.5 billion.³ More granularly, Marianne Lake, CEO of Consumer & Community Banking, has stated that technology investments within her division, which are heavily infused with AI, generate more than a two-times return on investment and consistently pay back within five years.⁵ Across the firm, investments specifically in AI and machine learning delivered a 35% increase in value in the last year alone, indicating an accelerating pace of returns.⁵ Cumulatively, it has been reported that AI initiatives have already saved the bank nearly $1.5 billion through the combined impact of fraud prevention, trading optimization, operational efficiencies, and improved credit decisions.⁸
由总裁兼首席运营官丹尼尔·平托预测的主要数据是，人工智能计划预计每年将为业务创造 15 亿至 20 亿美元的价值。 ³ 这项价值来自于收入增长、成本降低和成本避免的综合效应。该公司对这一回报的信心正在增强；这一预测是从之前的 10 亿至 15 亿美元估计值有所增加。 ³ 更具体地说，消费者与社区银行的首席执行官玛丽安·莱克表示，她部门内的技术投资，这些投资大量融入了人工智能，产生了超过两倍的回报率，并且始终在五年内收回成本。 ⁵ 在整个公司范围内，专门投资于人工智能和机器学习的项目仅在过去一年就实现了价值 35%的增长，表明回报速度正在加速。 ⁵ 累计来看，据报道，人工智能计划通过欺诈预防、交易优化、运营效率和信用决策改进的综合影响，已经为银行节省了近 15 亿美元。 ⁸

Efficiency and Productivity Gains: High-Impact Operational Returns
效率与生产力提升：高影响力的运营回报

Beyond direct financial returns, AI is generating profound operational leverage by automating manual processes, improving accuracy, and freeing up human capital for higher-value work.
除了直接的经济回报外，人工智能通过自动化手动流程、提高准确性和释放人力资本用于更高价值的工作，正在产生深刻的运营杠杆效应。

• Cost Avoidance: Perhaps the most powerful, yet often overlooked, ROI metric is defensive. The firm faces a 12% compound annual growth rate in fraud attacks, a threat that could lead to catastrophic escalating costs for an institution of its size. By using AI to hold the cost of fraud flat, JPMC is achieving a massive cost avoidance that directly protects its “fortress balance sheet” philosophy and saves hundreds of millions, if not billions, in potential losses.⁵
  成本避免：也许最强大但往往被忽视的投资回报指标是防御性的。该机构面临着 12%的复合年增长率欺诈攻击，这一威胁可能导致其规模的机构出现灾难性的成本不断上升。通过使用人工智能将欺诈成本稳定在当前水平，摩根大通正在实现巨大的成本避免，这直接保护了其“堡垒资产负债表”的理念，并可能节省数十亿甚至数百亿美元潜在的损失。 ⁵
• Labor Productivity: The efficiency gains in terms of work hours are staggering. The COiN platform’s automation of legal document review saves over 360,000 work hours annually.⁷ The firm-wide LLM Suite is saving its 200,000+ users “several hours per week” on routine tasks.⁵ In specialized roles, the impact is even more pronounced: investment bankers are automating 40% of their research tasks, and portfolio managers and traders have seen research time cut by as much as 83% using AI tools.²³
  劳动生产率：在工作时间方面的效率提升令人惊叹。COiN 平台的自动化法律文件审查每年节省超过 36 万小时的工作时间。 ⁷ 公司范围的 LLM 套件让 200,000 多名用户在常规任务上“每周节省数小时”。 ⁵ 在专业岗位上，影响更为显著：投资银行家正在自动化 40%的研究任务，而投资组合经理和交易员使用 AI 工具使研究时间减少了高达 83%。 ²³
• Process Improvement: AI is systematically rooting out inefficiencies in core processes. In anti-money laundering (AML) compliance, AI models have produced a 95% reduction in false positives, allowing human investigators to focus on legitimate risks.⁹ In payments, AI-powered validation has cut account rejection rates by 15-20%.³² Within the Global Payments division, AI models have cut manual exceptions by over 50%, delivering significant operating leverage even as transaction volumes surged by more than 50%.⁵
  流程改进：AI 正系统地消除核心流程中的低效环节。在反洗钱（AML）合规方面，AI 模型将误报率降低了 95%，使人类调查人员能专注于真实风险。 ⁹ 在支付领域，AI 驱动的验证将账户拒绝率降低了 15-20%。 ³² 在全球支付部门，AI 模型将手动例外情况减少了 50%以上，即使交易量激增超过 50%，也实现了显著的运营杠杆。 ⁵

The Revenue Generation Engine: AI’s Contribution to Growth
增长引擎：AI 对增长的贡献

JPMC’s AI strategy is not solely about cutting costs; it is also a potent engine for revenue generation, particularly in client-facing businesses.
摩根大通（JPMC）的 AI 战略不仅关乎削减成本，也是客户导向业务的强大收入增长引擎。

In the Asset & Wealth Management division, the deployment of GenAI-driven tools was a key contributor to a 20% year-over-year increase in gross sales between 2023 and 2024.⁸ Tools like Coach AI, which helps advisors access information 95% faster, are enabling them to provide quicker, more tailored investment advice, especially during periods of market turmoil. This capability has been directly credited with helping the firm boost sales and add new clients.⁸ The productivity gains are so significant that the firm projects these AI tools will empower advisors to expand their individual client rosters by as much as 50% over the next three to five years, a direct driver of future revenue growth.⁸
在资产管理与财富管理部门，GenAI 驱动工具的部署是 2023 年至 2024 年间毛销售额同比增长 20%的关键因素。 ⁸ 像 Coach AI 这样的工具，能帮助顾问以 95%的速度获取信息，使他们能够提供更快速、更个性化的投资建议，尤其是在市场动荡时期。这项能力直接促成了公司销售额的提升和新客户的增加。 ⁸ 生产力的提升非常显著，公司预计这些 AI 工具将在未来三到五年内使顾问的客户团队规模扩大 50%，这是未来收入增长的主要驱动力。 ⁸

The firm’s public disclosure of these ROI metrics is, in itself, a competitive strategy. It builds confidence among investors, justifying the massive technology budget. It also sets an exceptionally high bar for competitors, who may be pressured by their own stakeholders to disclose similar metrics that they may not be able to produce. Internally, this focus on quantifiable results fosters a culture of accountability, ensuring that AI projects are treated as business initiatives expected to deliver measurable value, not as open-ended science experiments.
该公司公开披露这些投资回报率指标本身就是一种竞争策略。它增强了投资者的信心，为庞大的技术预算提供了合理性。同时，它也为竞争对手设定了一个极其高的标准，这些竞争对手可能因自身利益相关者的压力而被迫披露类似的指标，但他们可能无法实现这些指标。在内部，这种对可量化结果的关注培养了问责文化，确保人工智能项目被视为预期能带来可衡量价值的商业计划，而不是无限制的科学实验。

Quantified AI Impact Across Business Lines
跨业务线的量化 AI 影响

Business Area  业务领域	Use Case / Platform  用例/平台	Impact Metric  影响指标	Quantified Value  量化价值	Source(s)  来源
Firm-wide  全公司范围	Overall AI Initiatives  整体人工智能计划	Annual Business Value  年度业务价值	$1.5B – $2.0B  1.5 亿 – 2 亿美元	3
Corporate (Legal)  企业（法律）	COiN Platform  COiN 平台	Labor Productivity  劳动生产率	360,000+ hours saved annually 每年节省超过 36 万小时	7
Corporate (Legal)  企业（法律）	COiN Platform  COiN 平台	Error Reduction  减少错误	~80% reduction in compliance errors ~80%减少合规错误	7
Corporate (Development)  企业（开发）	AI Coding Assistants  AI 编码助手	Development Velocity  开发速度	+70% code deployments in 2 years +70% 代码部署在 2 年内	5
CCB (Fraud)  CCB（欺诈）	AI Fraud Detection  AI 欺诈检测	Cost Avoidance  成本避免	Fraud costs held flat despite +12% CAGR in attacks 欺诈成本保持不变，尽管攻击增长率达 12%	5
CCB (Payments)  CCB（支付）	AI Payment Validation  AI 支付验证	Process Improvement  流程改进	15-20% reduction in account validation rejections 账户验证拒绝率降低 15-20%	32
CIB (Trading)  CIB（交易）	AI Research Tools  AI 研究工具	Research Productivity  研究生产力	Up to 83% reduction in research time 研究时间减少高达 83%	31
AWM (Sales)  AWM（销售）	GenAI-driven Tools  GenAI 驱动的工具	Revenue Generation  收入生成	+20% YoY increase in gross sales (2023-2024) +20%的年增长率（2023-2024）	8
AWM (Advisory)  顾问服务	Coach AI Tool  教练 AI 工具	Advisor Productivity  顾问生产力	Up to 95% faster information retrieval for advisors 顾问信息检索速度提升高达 95%	8
AWM (Advisory)  AWM（顾问）	Coach AI Tool  教练 AI 工具	Future Revenue Growth  未来收入增长	Projected 50% expansion of advisor client rosters 预计顾问客户团队将扩大 50%	8

The Human Element: Talent, Training, and Transformation
人性因素：人才、培训与转型

JPMorgan Chase’s AI strategy is as much about human capital as it is about technology. The firm is executing a sophisticated, dual-pronged approach: on one hand, it is waging an aggressive campaign to acquire and retain the world’s most elite AI talent; on the other, it is strategically using AI-driven efficiency to reshape its broader workforce and cost base. This “barbell” talent strategy is designed to create a more polarized but ultimately more productive organization, with a smaller, highly-paid group of AI creators enabling a larger, AI-augmented workforce.
摩根大通的人工智能战略不仅关乎技术，也关乎人才资本。该行正在执行一项复杂、双管齐下的策略：一方面，它正积极展开一场获取和留住全球顶尖人工智能人才的战役；另一方面，它正战略性地利用人工智能驱动的效率来重塑其整体员工队伍和成本结构。这种“哑铃式”人才战略旨在打造一个更加两极分化但最终更具生产力的组织，由一小部分高薪人工智能创造者带动一大群人工智能辅助的员工队伍。

The War for Talent: Acquiring the AI Elite
人才争夺战：获取人工智能精英

Recognizing that leadership in AI requires world-class expertise, JPMC has made talent acquisition a top priority. The firm has built a formidable internal team of over 2,000 AI and machine learning (ML) experts and data scientists, with an ambitious target to grow this specialized workforce to 5,000.¹⁰ This includes a 200-person elite AI Research lab, which functions like that of a major technology company, pursuing primary research and attracting top-tier academic and industry talent.⁴
认识到人工智能领域的领导力需要世界级的专业知识，摩根大通已将人才获取列为首要任务。该行已组建了一支由 2000 多名人工智能和机器学习（ML）专家及数据科学家组成的强大内部团队，并雄心勃勃地计划将这支专业队伍扩大到 5000 人。 ¹⁰ 这包括一个由 200 人组成的精英人工智能研究实验室，其运作方式类似于一家大型科技公司，从事基础研究并吸引顶尖的学术和行业人才。 ⁴

The firm is hiring aggressively to meet these goals. In mid-2024, it had over 75 open positions in AI and related fields.³³ To compete directly with big tech, JPMC offers highly competitive compensation packages, with roles such as an Applied AI ML Senior Associate commanding salaries of up to $195,000 annually.³³ The strategy is not just to buy talent but to attract it by offering a unique value proposition: the opportunity to work on some of the most complex and high-stakes problems in finance, powered by the firm’s unparalleled data assets and resources.¹³ As Chief Data and Analytics Officer Teresa Heitsenrether stated, “We recognize that our people are our strength and the diverse talents they bring to our global workforce are directly linked to our success”.³³
该机构正积极招聘以实现这些目标。2024 年年中，它在人工智能及相关领域有超过 75 个空缺职位。 ³³ 为了与大科技公司直接竞争，摩根大通提供极具竞争力的薪酬方案，例如应用人工智能机器学习高级副理的年薪高达 19.5 万美元。 ³³ 该战略不仅是为了购买人才，而是通过提供独特的价值主张来吸引人才：有机会在金融领域一些最复杂、最高风险的问题上工作，这些工作由该机构的无与伦比的数据资产和资源支持。 ¹³ 正如首席数据与分析官特蕾莎·海茨纳特雷尔所说：“我们认识到，我们的员工是我们的力量，他们为全球员工队伍带来的多样化才能直接关系到我们的成功”。 ³³

Attrition as a Friend: The Strategic Headcount Pause
人才流失作为朋友：战略性员工规模暂停

In a significant strategic pivot that signals immense confidence in its AI capabilities, JPMC’s leadership is actively managing its overall headcount. CFO Jeremy Barnum has publicly instructed managers to “resist head count growth where possible and increase their focus on efficiency”.³⁵ This directive is explicitly linked to the productivity gains being unlocked by AI.
在一场重大的战略转变中，这表明对人工智能能力的极大信心，摩根大通的管理层正在积极管理其整体员工规模。首席财务官杰里米·巴 NUM 公开指示管理者“尽可能抵制员工规模增长，并增加对效率的关注”。 ³⁵ 这一指令明确与人工智能带来的生产力提升相关联。

This is not a theoretical exercise. Marianne Lake, CEO of Consumer & Community Banking, has forecasted a 10% workforce reduction in operational areas such as account services and processing, calling that a “conservative estimate” and betting that AI will deliver even deeper cuts.³⁵ This aligns perfectly with CEO Jamie Dimon’s mantra that “attrition is your friend,” which underscores a clear strategy to leverage natural workforce turnover, augmented by AI-driven automation, to fundamentally reshape the firm’s cost structure.³⁵ Consequently, new hiring is being narrowly focused on “high-certainty,” directly revenue-generating roles, such as relationship bankers and financial advisors, who can be augmented by AI tools to become more productive.³⁵ This decision to pause broad hiring is a powerful, tangible bet that the firm’s AI strategy is a genuine driver of operating leverage, capable of absorbing future business growth without a corresponding increase in staff.
这不是一个理论练习。消费者与社区银行的首席执行官玛丽安·莱克预测，在账户服务和处理等运营领域将减少 10%的劳动力，称其为“保守的估计”，并赌人工智能将带来更大幅度的削减。 ³⁵ 这与首席执行官杰米·戴蒙的箴言“人员流失是你的朋友”完美契合，这突显了利用自然劳动力更替，并通过人工智能驱动的自动化来根本重塑公司成本结构的明确策略。 ³⁵ 因此，新的招聘将狭隘地集中在“高确定性”直接创收的职位上，例如关系银行家和财务顾问，他们可以通过人工智能工具来提高生产力。 ³⁵ 暂停大规模招聘的决定是一个强有力的、具体的赌注，即公司的 AI 战略是运营杠杆的真正驱动因素，能够在没有相应增加人员的情况下吸收未来的业务增长。

Reskilling for the AI Era: Investing in the Future Workforce
为人工智能时代而再培训：投资未来劳动力

JPMC recognizes that the workforce transformation driven by AI requires a massive investment in training and reskilling. The firm invests $300 million in employee training each year, supplemented by a $350 million global workforce investment dedicated to piloting new education programs and improving collaboration between employers and educational institutions.¹⁰
摩根大通认识到，由人工智能驱动的劳动力转型需要大量投资于培训和技能再培养。该公司每年投入 3 亿美元用于员工培训，同时辅以 3.5 亿美元全球劳动力投资，用于试点新的教育项目，并改善雇主与教育机构之间的合作。 ¹⁰

The firm has made a commitment to redeploy and reskill employees in roles impacted by automation, transitioning them to higher-value duties.³⁸ This training is becoming increasingly sophisticated and role-specific. For example, in a clear sign of how work is changing, new analyst hires in the Asset & Wealth Management group now receive mandatory training in prompt engineering to ensure they can effectively leverage generative AI tools like the LLM Suite in their daily workflows.¹⁵ The firm also runs a 15-week paid “ReEntry Program” to help experienced professionals who have been on a career break relaunch their careers, providing them with updated skills and on-the-job training.⁴⁰
该机构承诺重新部署和重新培训受自动化影响的员工，将他们转移到更高价值的职责上。 ³⁸ 这项培训正变得越来越复杂和角色特定。例如，作为一个明显的信号，工作正在如何变化，资产管理与财富管理团队的新分析师现在必须接受提示工程的强制培训，以确保他们能够在日常工作中有效利用生成式 AI 工具，如 LLM 套件。 ¹⁵ 该机构还运行一个为期 15 周的付费“重返计划”，以帮助那些中断职业生涯的资深专业人士重新启动他们的职业生涯，为他们提供更新技能和在职培训。 ⁴⁰

Beyond its own walls, JPMC invests in building a long-term talent pipeline. It has committed $5.3 million to the Talent Ready Initiative in Washington, D.C., to strengthen the IT talent pool and has partnered with local governments and community organizations in cities like Detroit and San Francisco to support workforce development and expand access to apprenticeships in fields like technology and finance.¹⁰
超越自身围墙，摩根大通投资于建立长期的人才管道。它在华盛顿特区承诺投入 530 万美元用于“人才准备计划”，以加强 IT 人才库，并与底特律和旧金山等城市的当地政府和社区组织合作，支持劳动力发展，扩大在技术和金融等领域的学徒制机会。 ¹⁰

The Competitive Arena: Benchmarking Against Wall Street Rivals
竞争领域：与华尔街竞争对手进行基准测试

JPMorgan Chase’s claim to AI dominance is not made in a vacuum. A critical assessment requires benchmarking its strategy, scale, and results against its primary Wall Street competitors: Bank of America, Goldman Sachs, and Morgan Stanley. This analysis reveals that while all major banks are investing heavily in AI, JPMC’s approach is differentiated by its sheer scale, its enterprise-wide platform philosophy, and its externally validated leadership.
摩根大通在人工智能领域的统治地位并非空穴来风。要做出关键性评估，需要将其战略、规模和成果与华尔街主要竞争对手——美国银行、高盛和摩根士丹利——进行对比。这项分析显示，尽管所有主要银行都在大力投资人工智能，但摩根大通的方法因其庞大的规模、企业级平台理念以及外部验证的领导力而与众不同。

The Evident AI Index: A Third-Party Validation of Leadership
显见人工智能指数：第三方领导力验证

The most direct external validation of JPMC’s leadership comes from the Evident AI Index, a public benchmark that assesses the AI maturity of the world’s largest banks. For three consecutive years, JPMorgan Chase has been ranked #1 for overall AI capabilities.⁹ The index evaluates institutions across four key pillars—Talent, Innovation, Leadership, and Trust—and JPMC’s consistent top ranking, particularly its #1 position in the Innovation and Transparency sub-pillars, provides strong evidence of a well-rounded and effective strategy.⁹
JPMC 领导力的最直接外部验证来自 Evident AI Index，这是一个评估全球最大银行 AI 成熟度的公共基准。连续三年，摩根大通在整体 AI 能力方面排名第一。该指数从人才、创新、领导力和信任四个关键支柱对机构进行评估，JPMC 持续排名第一，尤其是在创新和透明度子支柱中排名第一，这有力地证明了其全面而有效的战略。

The data behind the ranking reveals the scale of JPMC’s commitment. The firm possesses the largest AI workforce of all financial organizations analyzed by Evident, employing more AI researchers than the next seven largest contenders combined.²⁹ This concentration of talent creates a critical mass of expertise that enables the firm to tackle complex problems and accelerate innovation at a pace its rivals cannot easily match. This #1 ranking is more than a vanity metric; it is a powerful talent acquisition tool. For elite AI professionals seeking to work on cutting-edge problems with maximum impact, the publicly validated leadership position signals that JPMC is the premier destination in finance, creating a virtuous cycle where the ranking attracts top talent, who in turn produce the innovative work that solidifies the ranking.
排名背后的数据揭示了摩根大通承诺的规模。该公司在 Evident 分析的所有金融机构中拥有最大的 AI 员工队伍，其 AI 研究人员的数量超过了前七家最大竞争对手的总和。 ²⁹ 这种人才集中形成了关键的专业知识规模，使公司能够应对复杂问题，并以竞争对手难以轻易匹敌的速度加速创新。这个#1#排名不仅仅是一个虚荣指标；它是一个强大的人才招聘工具。对于寻求解决尖端问题并产生最大影响的精英 AI 专业人士来说，公开验证的领导地位表明摩根大通是金融领域的首选目的地，从而形成了一个良性循环：排名吸引顶尖人才，而这些人才反过来产生的创新工作又巩固了排名。

JPMC vs. Bank of America: A Battle of Strategic Philosophies
摩根大通与美国银行：战略哲学的较量

Bank of America (BofA) represents a formidable competitor, but its AI strategy differs from JPMC’s in philosophy and focus.
美国银行（BofA）是一个强大的竞争对手，但其人工智能战略在理念和重点上与摩根大通（JPMC）有所不同。

• Scale of Investment: While BofA’s technology spending is substantial at around $13 billion annually, its allocation to new initiatives including AI was reported as $4 billion in 2025.⁴³ This is a significant sum, but appears smaller in scale than JPMC’s overall technology budget of $18 billion and its broad-based investment approach.¹
  投资规模：虽然花旗集团的科技支出每年约为 130 亿美元，但在 2025 年，其用于新举措（包括人工智能）的分配据报道为 40 亿美元。 ⁴³ 这是一笔巨款，但与摩根大通整体 180 亿美元的科技预算及其广泛的投资策略相比，规模似乎显得较小。 ¹
• Strategic Focus: The primary difference lies in their strategic approach. JPMC is pursuing a broad, enterprise-wide platform and ecosystem strategy, building foundational tools like OmniAI and the LLM Suite to transform the entire firm from the inside out.⁴ BofA’s strategy appears more
  战略重点：主要区别在于他们的战略方法。摩根大通正在追求一个广泛的企业级平台和生态系统战略，通过构建像 OmniAI 和 LLM 套件这样的基础工具，从内部彻底转型整个公司。 ⁴ 花旗的策略似乎更
  product-centric, anchored by its flagship AI-powered virtual assistant, “Erica.” Launched in 2018, Erica has been highly successful, handling over 2.5 billion client interactions.⁴⁴ BofA’s approach focuses on scaling and expanding this successful application, rolling out “Erica for Employees” and leveraging the underlying technology for tools like “askMerrill”.⁴³ This is a strategy of building on a proven product, whereas JPMC’s is one of building a foundational factory for all future products.
  以产品为中心，由其旗舰 AI 虚拟助手“Erica”作为核心。自 2018 年推出以来，Erica 取得了巨大成功，处理了超过 25 亿次的客户互动。 ⁴⁴ 美国银行的方法侧重于扩展和推广这一成功应用，推出了“Erica for Employees”，并利用底层技术为工具如“askMerrill”赋能。 ⁴³ 这是一种基于成熟产品的策略，而摩根大通采用的是建立所有未来产品的基础工厂的策略。
• Patents: BofA has a very strong focus on intellectual property, holding over 1,200 patents related to AI and machine learning, which suggests a strategy of protecting specific innovations and applications.⁴³
  专利：美国银行非常重视知识产权，持有超过 1200 项与人工智能和机器学习相关的专利，这表明其保护特定创新和应用的战略。 ⁴³

JPMC vs. Goldman Sachs & Morgan Stanley: Scale vs. Niche
JPMC 与高盛和摩根士丹利：规模与专精

The competitive dynamic with Goldman Sachs and Morgan Stanley highlights a different strategic trade-off: JPMC’s industrial scale versus the focused, high-margin approach of its investment banking-centric rivals.
与高盛和摩根士丹利的竞争动态凸显了不同的战略权衡：JPMC 的工业规模与其投资银行中心竞争对手专注、高利润率的方法之间的权衡。

• Strategic Focus: The battle here is not over who can build better AI, but for what purpose. JPMC’s strategy is to build an “AI factory” that generates value through network effects, operational leverage, and enterprise-wide data synergies across its universal banking model. Its goal is to lower the marginal cost of the next AI application to near zero. In contrast, Goldman Sachs and Morgan Stanley are building “bespoke workshops” designed to create high-margin, specialized products for their core Investment Banking and Wealth Management franchises.⁴ They are playing a precision and margin game, while JPMC is playing a volume and scale game.
  战略重点：这里的竞争不是关于谁能构建更好的 AI，而是为了什么目的。JPMC 的战略是构建一个“AI 工厂”，通过网络效应、运营杠杆及其全渠道银行模式的企业级数据协同效应来创造价值。其目标是使下一个 AI 应用的边际成本接近于零。相比之下，高盛和摩根士丹利正在构建“定制工坊”，旨在为其核心投资银行和财富管理业务创建高利润率的专用产品。 ⁴ 他们玩的是精准和利润率的游戏，而 JPMC 玩的是数量和规模的游戏。
• Flagship Initiatives: This difference is clear in their flagship initiatives. JPMC’s is the enterprise-wide LLM Suite, deployed to over 200,000 employees to drive broad productivity gains.⁵ Goldman’s is the more focused
  旗舰计划：这种差异在他们的旗舰计划中很明显。JPMC 的是企业级 LLM 套件，部署给超过 20 万名员工以推动广泛的效率提升。 ⁵ 高盛的则更加专注
  GS AI Assistant, rolled out to its bankers and traders to support specific, high-value workflows.⁴⁷ Morgan Stanley has been a pioneer in this area, developing a generative AI assistant in partnership with OpenAI specifically for its
  GS AI 助手，已向其银行家和交易员推出，以支持特定的、高价值的业务流程。 ⁴⁷ 摩根士丹利在这一领域一直处于领先地位，与 OpenAI 合作开发了一个生成式 AI 助手，专门用于其
  16,000 wealth management financial advisors to help them access the firm’s vast intellectual capital more efficiently.⁴⁹
  16,000 名财富管理金融顾问将协助他们更高效地获取公司的庞大知识资本。 ⁴⁹
• Workforce Impact: The approaches to human-machine collaboration also differ. While JPMC is using AI to justify a broad pause on hiring in operational roles, Goldman Sachs is experimenting with fully autonomous AI coding agents like “Devin” to augment its elite engineering teams, signaling a deeper exploration of AI as a “digital employee” rather than just a tool.⁵²
  对员工的影响：人类与机器协作的方式也存在差异。虽然摩根大通正利用人工智能为其运营岗位的广泛招聘暂停辩护，高盛则正在试验完全自主的人工智能编码代理，如“德文”（Devin），以增强其精英工程团队，这表明其对人工智能作为“数字员工”而非仅仅是工具的探索更为深入。 ⁵²

Competitive AI Matrix: JPMC vs. Rivals
竞争 AI 矩阵：摩根大通与竞争对手

Key Strategic Pillar  关键战略支柱	JPMorgan Chase  摩根大通	Bank of America  美国银行	Goldman Sachs  高盛	Morgan Stanley  摩根士丹利
Scale of Investment  投资规模	$18B total tech budget (2025) 1 180 亿美元的总技术预算（2025 年） 1	$13B total tech budget; $4B for new AI/tech initiatives (2025) 44 $130 亿总技术预算；$40 亿用于新的 AI/技术计划（2025 年） 44	Significant, but not specified; focused on high-impact use cases 48 很大，但未具体说明；专注于高影响力的应用场景 48	$4.6B total ICT spend (2023); focused on wealth management 55 $46 亿总 ICT 支出（2023 年）；专注于财富管理 55
Primary Strategic Focus  主要战略重点	Enterprise-wide integration; proprietary platform building; fundamental R&D 企业级集成；专有平台建设；基础研发	Consumer experience; employee productivity; scaling the “Erica” product ecosystem 消费者体验；员工生产力；“Erica”产品生态系统的扩展	Niche, high-margin Investment Banking & Wealth Management applications; specialized copilots 专业、高利润的投资银行与财富管理应用；专用助手	Empowering financial advisors; enhancing wealth management client service 赋能财务顾问；提升财富管理客户服务
Flagship AI Initiative(s) 旗舰 AI 计划	LLM Suite (Internal, 200k+ users), OmniAI (Integration), COiN (Ops) 5 LLM 套件（内部，20 万+用户），OmniAI（集成），COiN（运营） 5	Erica (Customer), Erica for Employees (Internal) 43 Erica（客户），Erica for Employees（内部） 43	GS AI Assistant (Internal, role-specific), “Devin” AI agent testing 47 GS AI 助手（内部，角色特定），“Devin” AI 代理测试 47	AI @ Morgan Stanley Assistant (for Financial Advisors), AskResearchGPT 49 AI @ Morgan Stanley 助手（面向金融顾问），AskResearchGPT 49
Stated Workforce Impact  声明的工作 force 影响	“Resist headcount growth”; 10% reduction in some ops roles; “attrition is your friend” 35 “抵制员工数量增长”；部分运营岗位削减 10%；“人员流失是你的朋友” 35	Redeploying staff freed from rote tasks into higher-value, AI-building roles 45 将从事重复性工作的人员重新部署到更高价值、构建 AI 的岗位 45	Augmenting engineering workforce with AI agents; enhancing developer productivity 48 通过 AI 代理增强工程团队；提升开发者生产力 48	Enhancing advisor efficiency; enabling them to serve more clients 49 提升顾问效率；使他们能够服务更多客户 49
Evident AI Index Rank  明显的 AI 指数排名	#1 9	#15 56	#4 56	Not in top 5  不在前五

Navigating the Future: Risks, Ethics, and the Quantum Frontier
引领未来：风险、伦理与量子前沿

While JPMorgan Chase has established a formidable lead in the AI race, sustaining this dominance requires navigating a complex and evolving landscape of risks, ethical considerations, and technological disruption. The firm’s strategy demonstrates a sophisticated understanding of these future challenges, focusing on building a proactive governance framework that doubles as a competitive moat and making long-term bets on next-generation technologies like quantum computing.
尽管摩根大通在人工智能竞赛中已确立了强大的领先地位，但要维持这一优势需要应对复杂且不断变化的风险、伦理考量和技术颠覆。该公司的战略展现了其对这些未来挑战的深刻理解，重点在于构建一个既能作为竞争优势壁垒，又能主动进行治理的框架，并对量子计算等下一代技术进行长期投资。

Governance as a Moat: Weaponizing Compliance
以治理为护城河：将合规武器化

In a highly regulated industry like finance, a single major incident of algorithmic bias in lending or a significant data privacy breach could result in billions of dollars in fines, severe reputational damage, and restrictive regulatory consent decrees. JPMC’s investment in “Ethical AI” is therefore not an act of altruism, but a pragmatic risk management strategy designed to build a durable competitive advantage. By building a best-in-class governance framework, the firm is de-risking its AI-driven future, which will allow it to deploy AI into more sensitive, high-value areas faster and with greater confidence than competitors who are still grappling with the basics.
在高度监管的金融行业，借贷算法中的单一重大偏见事件或重大数据隐私泄露可能导致数十亿美元的罚款、严重的声誉损害以及限制性的监管许可裁决。摩根大通投资“道德 AI”因此并非利他行为，而是一种务实的风险管理策略，旨在建立持久的竞争优势。通过构建一流的治理框架，该机构正在降低其 AI 驱动的未来风险，这将使其能够比仍在应对基础问题的竞争对手更快、更有信心地将 AI 部署到更敏感、高价值领域。

The firm has established a robust internal governance structure, including an Explainable AI Center of Excellence, a dedicated firm-wide Model Risk Governance function, and an AI governance committee to oversee the development and deployment of all models.⁴ This structure ensures that concepts like Responsible AI and Ethical AI are not optional but are required for any application that leverages customer data.⁵⁸
该机构已建立完善的内部治理结构，包括可解释 AI 卓越中心、专门的全公司模型风险治理职能以及负责监督所有模型开发和部署的 AI 治理委员会。 ⁴ 这种结构确保“负责任的 AI”和“道德 AI”等概念并非可选，而是任何利用客户数据的应用程序都必须满足的要求。 ⁵⁸

JPMC is proactively addressing specific ethical risks. It has patented an algorithmic bias evaluation framework designed to audit risk assessment models, such as those used in lending, to ensure they do not create disparate impacts on protected groups.²⁶ To combat the risk of training models on biased historical data, the firm is a leader in the use of synthetic data. By generating realistic, artificial datasets, JPMC can train its models on more equitable and complete scenarios without exposing sensitive client information, a practice that also strengthens compliance with emerging regulatory guidelines.²⁶
摩根大通主动应对特定的伦理风险。该公司已获得一项算法偏见评估框架的专利，该框架旨在审计风险评估模型（例如用于贷款的模型），以确保它们不会对受保护群体产生差异化影响。 ²⁶ 为了应对在存在偏见的历史数据上训练模型的风险，该公司是合成数据使用的领先者。通过生成逼真的、人造的数据集，摩根大通可以在更公平和完整的场景下训练其模型，而无需暴露敏感的客户信息，这一做法也有助于加强其对新兴监管指南的合规性。 ²⁶

Critically, JPMC is extending these high standards to its entire ecosystem. In a notable open letter, the firm signaled it is now requiring its third-party technology suppliers to demonstrate responsible AI practices, including providing detailed documentation on training data, model development, and fairness assessments.⁵⁹ This effectively weaponizes compliance, raising the security and ethical bar for its partners and creating a competitive moat against less rigorous rivals who may be exposed to vulnerabilities from their supply chain.⁴
关键的是，摩根大通正将其高标准扩展到整个生态系统。在一封显著的公开信中，该公司表明现在要求其第三方技术供应商展示负责任的 AI 实践，包括提供关于训练数据、模型开发和公平性评估的详细文档。 ⁵⁹ 这实际上将合规性武器化，提高了其合作伙伴的安全和道德标准，并为其创造了竞争优势，使其能够对抗那些可能因供应链问题而暴露于漏洞的较不严格的竞争对手。 ⁴

Implementation Headwinds: The Challenges of Scale
实施阻力：规模挑战

Despite its progress, JPMC’s path is not without significant challenges, primarily stemming from the sheer scale and complexity of its operations.
尽管取得了进展，摩根大通的道路并非没有重大挑战，主要源于其运营的巨大规模和复杂性。

• Technical Debt: The firm still operates over 100,000 legacy IT applications and databases.²⁷ Integrating advanced neural networks and deep learning algorithms with these dated systems is an immense challenge that requires extensive IT resources and specialized expertise. The process of integration, testing, and deployment can stretch for months for a single use case.²⁷
  技术债务：该公司的 IT 系统仍运行着超过 10 万个遗留应用程序和数据库。 ²⁷ 将先进的神经网络和深度学习算法与这些老旧系统集成是一个巨大的挑战，需要大量的 IT 资源和专业知识。集成、测试和部署单个用例的过程可能持续数月。 ²⁷
• Data Security & Privacy: In the age of generative AI, protecting sensitive financial data is paramount. The risk of data leakage has led the firm to prohibit the use of consumer AI chatbots like ChatGPT for work purposes, which necessitated the costly but necessary development of proprietary, secure tools like the LLM Suite.⁶¹ As the firm’s CISO has highlighted, many organizations are deploying AI systems they “fundamentally don’t understand,” creating security vulnerabilities that JPMC is working diligently to avoid.⁶⁰
  数据安全与隐私：在生成式 AI 时代，保护敏感的金融数据至关重要。数据泄露的风险导致公司禁止将 ChatGPT 等消费者 AI 聊天机器人用于工作目的，这促使公司开发成本高昂但必要的专有安全工具，如 LLM 套件。 ⁶¹ 正如公司首席信息安全官所强调的，许多组织正在部署他们“根本不理解”的 AI 系统，从而创造了安全漏洞，而 JPMC 正努力避免这种情况。 ⁶⁰
• Organizational Change: Perhaps the greatest challenge is cultural. Effectively incorporating AI into the day-to-day activities of a global workforce of over 300,000 people is a monumental task.⁸ As one analysis noted, the demand to build proofs of concept is often outstripping “the mechanics of the organisation to deliver the change”.³² Managing this transformation, reskilling the workforce, and adapting to evolving user behaviors and expectations are ongoing, complex challenges.⁶²
  组织变革：或许最大的挑战在于文化。将 AI 有效融入超过 30 万人的全球员工的日常活动中是一项艰巨的任务。 ⁸ 正如一项分析指出，构建概念验证的需求往往超过了“组织交付变革的机制”。 ³² 管理这场转型、重新培训员工、以及适应不断变化用户行为和期望都是持续且复杂的挑战。 ⁶²

Beyond AI: The Quantum Computing Bet
超越 AI：量子计算的赌注

JPMC’s long-term strategy extends beyond the current paradigm of AI. The firm is making a significant, forward-looking bet on quantum computing, a technology that promises to solve certain types of problems that are intractable for even the most powerful classical supercomputers and AI. This investment is a two-sided, long-term hedge against technological disruption. Defensively, it is preparing for “Q-Day”—the day quantum computers become powerful enough to break the cryptographic security that underpins the entire global financial system. Offensively, it is positioning itself to be the first to harness quantum’s power for financial modeling, optimization, and next-generation AI.
摩根大通（JPMC）的长期战略超越了当前的人工智能范式。该公司正对量子计算进行重大且具有前瞻性的投资，这项技术承诺解决那些连最强大的经典超级计算机和人工智能都无法处理的特定类型问题。这项投资是一种双向的长期对冲，以应对技术颠覆。防御性地，它正在为“量子计算大日”（Q-Day）做准备——即量子计算机强大到足以破解支撑整个全球金融体系加密安全的那一天。进攻性地，它正将自己定位为第一个利用量子力量进行金融建模、优化和下一代人工智能的公司。

JPMC was one of the first financial institutions globally to invest in quantum computing, building an internal team of scientists to explore its application to finance, AI, and cryptography.⁶³ The firm’s research focuses on two key areas:
摩根大通是全球最早投资量子计算领域的金融机构之一，组建了内部科学家团队，探索其在金融、人工智能和密码学领域的应用。 ⁶³ 该公司的研发重点集中在两个关键领域：

1. Quantum Algorithms: The team is developing novel quantum algorithms for use cases such as portfolio optimization, option pricing, and risk analysis.⁶³
   量子算法：团队正在开发用于投资组合优化、期权定价和风险分析等应用场景的新型量子算法。 ⁶³
2. Quantum-Resistant Security: JPMC is actively working on Quantum Key Distribution (QKD) networks, a form of quantum-secured communication that is mathematically proven to be resistant to attacks from future quantum computers. In a groundbreaking collaboration with Toshiba and Ciena, the firm demonstrated the first viable QKD network for securing a mission-critical blockchain application.⁶⁴
   抗量子安全：摩根大通正积极研发量子密钥分发（QKD）网络，这是一种经过数学证明能够抵抗未来量子计算机攻击的量子安全通信方式。通过与东芝和思科的开创性合作，该公司展示了首个可用于保障关键区块链应用的可行 QKD 网络。 ⁶⁴

In a major scientific breakthrough published in the journal Nature, JPMC researchers, in partnership with Quantinuum, Argonne National Laboratory, and Oak Ridge National Laboratory, demonstrated the first use of a quantum computer to generate “Certified Randomness”.⁶⁵ This achievement, which uses a quantum computer to produce random numbers that are provably unpredictable, is a vital milestone with critical applications in cryptography, statistical sampling, and complex simulations. This research places JPMC at the absolute frontier of a technology that will likely define the next generation of computing, ensuring its dominance can be sustained long after the current AI hype cycle has passed.
在《自然》杂志发表的一项重大科学突破中，JPMC 研究人员与 Quantinuum、阿贡国家实验室和橡树岭国家实验室合作，展示了首次使用量子计算机生成“认证随机性”。 ⁶⁵ 这一成就利用量子计算机生成可证明不可预测的随机数，是密码学、统计抽样和复杂模拟领域的重要里程碑。这项研究将 JPMC 置于定义下一代计算技术绝对前沿的位置，确保其主导地位能在当前人工智能热潮周期结束后长期维持。

Strategic Insights and Recommendations
战略洞察与建议

The comprehensive analysis of JPMorgan Chase’s artificial intelligence strategy reveals a multi-layered and deeply integrated approach that is actively reshaping the competitive landscape of financial services. The firm’s current dominance is not an accident of its size but the deliberate outcome of a cohesive vision. The following insights synthesize the key success factors and offer recommendations for competitors, investors, and the broader industry.
对 JPMorgan Chase 人工智能战略的全面分析揭示了一种多层次且深度融合的方法，正在积极重塑金融服务领域的竞争格局。该公司的当前主导地位并非其规模的偶然结果，而是统一愿景的刻意成果。以下洞察综合了关键成功因素，并为竞争对手、投资者和更广泛的行业提供建议。

The Blueprint for Dominance: A Synthesis of Key Success Factors
主导地位蓝图：关键成功因素的综合

JPMorgan Chase’s AI leadership is built upon a synergistic combination of strategic pillars that create a powerful, compounding advantage. Any organization seeking to understand or replicate this success must recognize the interplay of these five core elements:
摩根大通的人工智能领导地位建立在战略支柱的协同组合之上，形成了强大的、复合的优势。任何希望理解或复制这一成功的组织都必须认识到这五个核心要素之间的相互作用：

1. Unambiguous C-Suite Mandate: The strategy begins with a clear, consistent, and urgent directive from the CEO, which frames AI as a transformational force and aligns the entire organization behind it.
   明确的 C 级管理层指令：该战略始于 CEO 发出的清晰、一致且紧迫的指令，将人工智能定位为变革性力量，并使整个组织围绕其展开。
2. Massive and Sustained Investment: A technology budget of $18 billion, with billions explicitly targeted at AI, provides the necessary fuel. This spending has shifted from modernizing legacy systems to funding offensive, innovation-driven AI initiatives.
   巨额且持续的投入：一项价值 180 亿美元的科技预算，其中数十亿美元明确用于人工智能，为发展提供了必要的燃料。这项支出已从现代化遗留系统转向资助进攻性、创新驱动的 AI 计划。
3. A Foundational Data and Platform Strategy: The firm’s unparalleled proprietary data, combined with a modernized cloud infrastructure (JADE) and scalable in-house platforms (OmniAI, LLM Suite), creates an “AI factory” that lowers the marginal cost of new applications and accelerates deployment.
   基础数据和平台战略：公司无与伦比的专有数据，结合现代化的云基础设施（JADE）和可扩展的内部平台（OmniAI、LLM 套件），创建了一个“AI 工厂”，降低了新应用的机会成本，并加速了部署。
4. Enterprise-Wide Deployment with Quantifiable ROI: With over 600 use cases in production delivering $1.5-$2.0 billion in annual value, the firm has moved from theory to execution, embedding AI across every business line and fostering a culture of accountability focused on measurable returns.
   企业范围的部署与可量化的 ROI：已有超过 600 个用例在生产中运行，每年创造 15 亿至 20 亿美元的价值，公司已从理论走向实践，将 AI 嵌入到每个业务领域，并培养了一种以可衡量回报为焦点的问责文化。
5. Proactive Governance-as-a-Moat: By investing heavily in ethical AI, bias mitigation, and security, and by extending these high standards to its suppliers, JPMC turns a potential liability and regulatory burden into a strategic enabler that builds trust and allows for faster deployment into sensitive areas.
   主动的“治理即护城河”：通过大力投资伦理 AI、偏见缓解和安全性，并将这些高标准扩展到其供应商，摩根大通将潜在的负债和监管负担转化为战略推动者，建立信任并允许更快地部署到敏感领域。

Future Growth Trajectories and Emerging Battlegrounds
未来增长轨迹与新兴战场

While JPMC’s current strategy is heavily focused on driving internal productivity and efficiency, the next phase of value creation and competition will likely shift to new frontiers.
尽管摩根大通当前的战略主要集中于提升内部生产力和效率，但下一阶段的价值创造和竞争可能将转向新的领域。

• Client-Facing Generative AI: The firm has been cautious about deploying generative AI in direct, unmediated customer interactions due to risks of misinformation.⁶⁸ However, the next major battleground will be in this arena. The internal debate over potentially releasing powerful advisor tools like
  面向客户的生成式 AI：由于错误信息的风险，该公司在直接、未经中介的客户互动中部署生成式 AI 方面一直持谨慎态度。 ⁶⁸ 然而，下一个主要战场将出现在这个领域。关于可能发布强大的顾问工具等内部争论
  Connect Coach externally will be a key indicator of the firm’s evolving risk appetite and its ambition to lead in AI-driven client experience.⁶⁹
  通过外部连接教练将成为衡量该公司不断变化的风险偏好及其在人工智能驱动客户体验领域领导雄心的关键指标。 ⁶⁹
• The Rise of Agentic AI: The future of AI in finance lies not just in assisting with tasks but in automating entire workflows. The development of more advanced “agentic AI” capable of autonomous reasoning and execution is the next logical step. JPMC’s plans to develop AI agents with human-like reasoning for tasks like investment analysis will transform the nature of work for its employees and the services it offers to clients.²³
  代理型 AI 的崛起：金融领域 AI 的未来不仅在于辅助任务，更在于自动化整个工作流程。开发更高级的“代理型 AI”，使其具备自主推理和执行能力，是下一步合乎逻辑的发展。摩根大通计划开发具有类人推理能力的 AI 代理，用于投资分析等任务，这将改变其员工的工作性质以及向客户提供的服务的本质。 ²³
• The Monetization of Data: JPMC’s massive investment in making its data “AI-ready” has created an immensely valuable asset. The firm’s reported plans to begin charging FinTechs and data aggregators for access to its consumer bank data signals a potential new, high-margin revenue stream.⁷⁰ This would represent a strategic pivot from using data to improve internal operations to monetizing the data infrastructure itself as a service.
  数据的变现：摩根大通在使其数据“AI 化”方面的大量投资，创造了一项极具价值的资产。该机构报告计划开始向金融科技公司和数据聚合商收取其消费者银行数据的访问权限，这预示着一个潜在的新高利润收入来源。 ⁷⁰ 这将代表一种战略转变，从利用数据改进内部运营转变为将数据基础设施本身作为服务进行变现。

Recommendations for Competitors and Observers
对竞争对手和观察者的建议

Based on this analysis, several strategic recommendations emerge for different stakeholders.
基于这项分析，针对不同利益相关者，可以提出以下战略建议。

• For Competitors: A frontal assault on JPMC’s scale is likely a futile and capital-intensive endeavor. A more viable strategy is to focus on building deep, defensible niches where specialized expertise, superior customer intimacy, or a more agile culture can outperform JPMC’s industrial-scale approach. The non-negotiable price of entry into the AI race, however, is a comprehensive data modernization strategy; without clean, connected, and accessible data, no amount of algorithmic sophistication can succeed.
  对于竞争对手：直接挑战摩根大通（JPMC）的规模很可能是一项徒劳且资本密集的努力。更可行的策略是专注于建立深度防御的利基市场，在这些市场中，专业领域的专长、卓越的客户亲密关系或更敏捷的文化能够超越摩根大通的大规模工业化方法。然而，进入人工智能竞赛的不可协商的价格是一个全面的数据现代化战略；没有干净、互联和可访问的数据，再复杂的算法也无法成功。
• For Investors: It is crucial to look beyond headline spending figures and AI announcements. The key metrics to track when evaluating JPMC and its competitors are tangible evidence of execution and return. These include AI adoption rates (e.g., the growth in LLM Suite users), quantified ROI disclosures (e.g., the annual business value generated), and concrete progress in workforce reshaping (e.g., shifts in headcount between operational and revenue-generating roles). JPMC’s ability to navigate the significant long-term cultural challenges of its “barbell” talent strategy will be a critical determinant of its sustained success and profitability.
  对于投资者而言：关键在于超越头版头条的支出数字和人工智能宣告。在评估摩根大通及其竞争对手时，需要关注的是执行和回报的实体证据。这些指标包括人工智能采用率（例如，LLM 套件用户的增长）、量化的投资回报率披露（例如，每年产生的业务价值），以及劳动力重塑的具体进展（例如，运营角色和创收角色之间员工人数的变动）。摩根大通能否应对其“哑铃式”人才战略带来的重大长期文化挑战，将决定其持续成功和盈利能力的核心。
• For the Industry: JPMorgan Chase is actively setting the de facto standard for AI governance, security, and ethics in the financial sector. Its actions, such as requiring transparency and responsible AI practices from its third-party suppliers, will have a significant ripple effect.⁵⁹ Smaller firms and technology vendors should anticipate that these higher standards will become the baseline expectation across the industry. Proactively investing in robust governance and security will not just be a matter of compliance, but a prerequisite for doing business with the sector’s most powerful players.
  针对行业：摩根大通正积极为金融行业的 AI 治理、安全和伦理设定事实上的标准。其要求第三方供应商提高透明度和负责任的 AI 实践等行动将产生显著的影响。 ⁵⁹ 小型企业和科技供应商应预期这些更高的标准将成为整个行业的基准预期。积极投资于健全的治理和安全不仅将是合规的问题，更是与该行业最具影响力的参与者进行业务往来的先决条件。

Works cited  参考文献

1. JPMorganChase Makes Data “AI Ready” – Markets Media, accessed July 16, 2025, https://www.marketsmedia.com/jpmorganchase-makes-data-ai-ready/
   摩根大通将数据“AI 化” – Markets Media，2025 年 7 月 16 日访问，https://www.marketsmedia.com/jpmorganchase-makes-data-ai-ready/
2. AI gets a seat at the big table at JP Morgan Chase – Industry News | The Financial Express, accessed July 16, 2025, https://www.financialexpress.com/business/industry-ai-gets-a-seat-at-the-big-table-at-jp-morgan-chase-3877873/
   AI 在摩根大通的重要会议中占有一席之地——行业新闻 | 财经快报，2025 年 7 月 16 日访问，https://www.financialexpress.com/business/industry-ai-gets-a-seat-at-the-big-table-at-jp-morgan-chase-3877873/
3. JPMorgan Chase to equip 140K workers with generative AI tool – CIO Dive, accessed July 16, 2025, https://www.ciodive.com/news/JPMorgan-Chase-LLM-Suite-generative-ai-employee-tool/726772/
   摩根大通将为 14 万名员工配备生成式 AI 工具——CIO Dive，2025 年 7 月 16 日访问，https://www.ciodive.com/news/JPMorgan-Chase-LLM-Suite-generative-ai-employee-tool/726772/
4. JPMorgan Chase AI : Banking on AI Dominance – Klover.ai, accessed July 16, 2025, https://www.klover.ai/jpmorgan-chase-ai-banking-on-ai-dominance/
   JPMorgan Chase AI：押注人工智能主导地位 – Klover.ai，访问于 2025 年 7 月 16 日，https://www.klover.ai/jpmorgan-chase-ai-banking-on-ai-dominance/
5. JPMorgan Chase’s IT, AI bets: Where the returns are – Constellation Research, accessed July 16, 2025, https://www.constellationr.com/blog-news/insights/jpmorgan-chase-s-it-ai-bets-where-returns-are
   JPMorgan Chase 的 IT、人工智能投资：回报所在之处 – Constellation Research，访问于 2025 年 7 月 16 日，https://www.constellationr.com/blog-news/insights/jpmorgan-chase-s-it-ai-bets-where-returns-are
6. Omni Means “All” – J.P. Morgan, accessed July 16, 2025, https://www.jpmorgan.com/technology/news/omni-ai
   全意即“全部” – J.P. Morgan，访问于 2025 年 7 月 16 日，https://www.jpmorgan.com/technology/news/omni-ai
7. How JPMorgan Uses AI to Save 360,000 Legal Hours a Year | by …, accessed July 16, 2025, https://medium.com/@arahmedraza/how-jpmorgan-uses-ai-to-save-360-000-legal-hours-a-year-6e94d58a557b
   JPMorgan 如何利用人工智能每年节省 36 万小时法律工作 | by …，访问于 2025 年 7 月 16 日，https://medium.com/@arahmedraza/how-jpmorgan-uses-ai-to-save-360-000-legal-hours-a-year-6e94d58a557b
8. JPMorgan says AI helped boost sales, add clients in market turmoil – ET CIO, accessed July 16, 2025, https://cio.economictimes.indiatimes.com/news/artificial-intelligence/jpmorgan-says-ai-helped-boost-sales-add-clients-in-market-turmoil/120918406
   摩根大通表示 AI 助力提升销售额、增加客户——ET CIO，2025 年 7 月 16 日访问，https://cio.economictimes.indiatimes.com/news/artificial-intelligence/jpmorgan-says-ai-helped-boost-sales-add-clients-in-market-turmoil/120918406
9. Twimbit AI Spotlight: – JP Morgan Chase, accessed July 16, 2025, https://cdn.twimbit.com/uploads/2024/12/10151259/Twimbit-AI-Spotlight-JP-Morgan-1.pdf
   Twimbit AI 聚焦：——摩根大通，2025 年 7 月 16 日访问，https://cdn.twimbit.com/uploads/2024/12/10151259/Twimbit-AI-Spotlight-JP-Morgan-1.pdf
10. How JPMorgan Chase is preparing the workforce for the future of AI, accessed July 16, 2025, https://www.jpmorganchase.com/newsroom/stories/how-jpmc-is-preparing-workforce-for-ai
    摩根大通如何为 AI 的未来准备劳动力，2025 年 7 月 16 日访问，https://www.jpmorganchase.com/newsroom/stories/how-jpmc-is-preparing-workforce-for-ai
11. 2023 INVESTOR DAY – GLOBAL TECHNOLOGY – JPMorgan Chase, accessed July 16, 2025, https://www.jpmorganchase.com/content/dam/jpmc/jpmorgan-chase-and-co/investor-relations/documents/events/2023/jpmc-investor-day-2023/JPM-Investor-Day-2023-Final-Transcript_Global-Technology.pdf
    2023 投资者日——全球科技——摩根大通，2025 年 7 月 16 日访问，https://www.jpmorganchase.com/content/dam/jpmc/jpmorgan-chase-and-co/investor-relations/documents/events/2023/jpmc-investor-day-2023/JPM-Investor-Day-2023-Final-Transcript_Global-Technology.pdf
12. JPMorgan Chase: Digital transformation, AI and data strategy sets up generative AI, accessed July 16, 2025, https://www.constellationr.com/blog-news/insights/jpmorgan-chase-digital-transformation-ai-and-data-strategy-sets-generative-ai
    摩根大通：数字化转型、人工智能与数据战略推动生成式人工智能，访问于 2025 年 7 月 16 日，https://www.constellationr.com/blog-news/insights/jpmorgan-chase-digital-transformation-ai-and-data-strategy-sets-generative-ai
13. About – J.P. Morgan, accessed July 16, 2025, https://www.jpmorgan.com/technology/artificial-intelligence/about
    关于 – 摩根大通，访问于 2025 年 7 月 16 日，https://www.jpmorgan.com/technology/artificial-intelligence/about
14. Full Investor Day 2025 Presentation – JPMorgan Chase, accessed July 16, 2025, https://www.jpmorganchase.com/content/dam/jpmc/jpmorgan-chase-and-co/investor-relations/documents/events/2025/jpmc-2025-investor-day/full-presentation.pdf
    2025 年完整投资者日演示文稿 – 摩根大通，访问于 2025 年 7 月 16 日，https://www.jpmorganchase.com/content/dam/jpmc/jpmorgan-chase-and-co/investor-relations/documents/events/2025/jpmc-2025-investor-day/full-presentation.pdf
15. Case Study: JPMorgan is Setting the Standard for AI Adoption in …, accessed July 16, 2025, https://aiexpert.network/jpmorgan-ai/
    案例研究：摩根大通正在为…设定人工智能采用标准，访问于 2025 年 7 月 16 日，https://aiexpert.network/jpmorgan-ai/
16. JPMorgan Chase & Co. is furnishing this revised copy of … – SEC.gov, accessed July 16, 2025, https://www.sec.gov/Archives/edgar/data/19617/000001961725000329/annualreport-2024.pdf
    摩根大通公司正在提供这份修订版的…… – SEC.gov，访问日期为 2025 年 7 月 16 日，https://www.sec.gov/Archives/edgar/data/19617/000001961725000329/annualreport-2024.pdf
17. Full Investor Day 2024 Presentation – JPMorgan Chase, accessed July 16, 2025, https://www.jpmorganchase.com/content/dam/jpmc/jpmorgan-chase-and-co/investor-relations/documents/events/2024/jpmc-2024-investor-day/full-presentation.pdf
    2024 年完整投资者日演示文稿 – 摩根大通，访问日期为 2025 年 7 月 16 日，https://www.jpmorganchase.com/content/dam/jpmc/jpmorgan-chase-and-co/investor-relations/documents/events/2024/jpmc-2024-investor-day/full-presentation.pdf
18. Jamie Dimon Chairman and CEO of JPMorgan Chase at Data + AI Summit – YouTube, accessed July 16, 2025, https://www.youtube.com/watch?v=0hT-FJ7Ay4Y
    杰米·戴蒙作为摩根大通董事长兼首席执行官在数据+人工智能峰会上的发言 – YouTube，访问日期为 2025 年 7 月 16 日，https://www.youtube.com/watch?v=0hT-FJ7Ay4Y
19. How JPMorgan Chase’s infrastructure chief keeps the AI engine humming – CIO Dive, accessed July 16, 2025, https://www.ciodive.com/news/jpmorgan-chase-infrastructure-cio-ai-compute-strategy/738662/
    摩根大通基础设施负责人如何保持人工智能引擎高效运转 – CIO Dive，访问日期为 2025 年 7 月 16 日，https://www.ciodive.com/news/jpmorgan-chase-infrastructure-cio-ai-compute-strategy/738662/
20. Artificial Intelligence – Chase Alumni Association, accessed July 16, 2025, https://www.chasealum.org/article.html?aid=1692
    人工智能——Chase 校友协会，访问于 2025 年 7 月 16 日，https://www.chasealum.org/article.html?aid=1692
21. In the News – J.P. Morgan, accessed July 16, 2025, https://www.jpmorgan.com/technology/applied-ai-and-ml/machine-learning/news
    新闻动态——摩根大通，访问于 2025 年 7 月 16 日，https://www.jpmorgan.com/technology/applied-ai-and-ml/machine-learning/news
22. 10 ways JP Morgan is using AI [In Depth Case Study][2025] – DigitalDefynd, accessed July 16, 2025, https://digitaldefynd.com/IQ/jp-morgan-using-ai-case-study/
    摩根大通使用人工智能的 10 种方式 [深度案例研究][2025]——DigitalDefynd，访问于 2025 年 7 月 16 日，https://digitaldefynd.com/IQ/jp-morgan-using-ai-case-study/
23. Case Study: How JPMorgan Chase is Revolutionizing Banking Through AI, accessed July 16, 2025, https://aiexpert.network/ai-at-jpmorgan/
    案例研究：摩根大通如何通过人工智能革新银行业务，访问于 2025 年 7 月 16 日，https://aiexpert.network/ai-at-jpmorgan/
24. JPM Virtual Assistant | J.P. Morgan, accessed July 16, 2025, https://www.jpmorgan.com/payments/solutions/access/jpm-virtual-assistant
    JPM 虚拟助手 | 摩根大通，访问于 2025 年 7 月 16 日，https://www.jpmorgan.com/payments/solutions/access/jpm-virtual-assistant
25. How JP Morgan Built An AI Agent for Investment Research with LangGraph | LangChain Interrupt – YouTube, accessed July 16, 2025, https://www.youtube.com/watch?v=yMalr0jiOAc
    摩根大通如何使用 LangGraph 构建投资研究 AI 代理 | LangChain 中断 – YouTube，访问于 2025 年 7 月 16 日，https://www.youtube.com/watch?v=yMalr0jiOAc
26. JPMorgan Chase: Pioneering Ethical AI in Finance Through Synthetic Data and Bias Mitigation – AInvest, accessed July 16, 2025, https://www.ainvest.com/news/jpmorgan-chase-pioneering-ethical-ai-finance-synthetic-data-bias-mitigation-2506/
    摩根大通：通过合成数据和偏差缓解在金融领域开创道德 AI – AInvest，访问于 2025 年 7 月 16 日，https://www.ainvest.com/news/jpmorgan-chase-pioneering-ethical-ai-finance-synthetic-data-bias-mitigation-2506/
27. “Transforming Financial Services: The Impact of AI on JP Morgan Chase’s Operational Efficiency and Decision- Making”, accessed July 16, 2025, https://ijsret.com/wp-content/uploads/2024/01/IJSRET_V10_issue1_138.pdf
    “变革金融服务：AI 对摩根大通运营效率和决策的影响”，访问于 2025 年 7 月 16 日，https://ijsret.com/wp-content/uploads/2024/01/IJSRET_V10_issue1_138.pdf
28. JPMorgan says it’s using AI for all these things once done by humans – eFinancialCareers, accessed July 16, 2025, https://www.efinancialcareers.com/news/jpmorgan-ai
    摩根大通表示，它正在使用人工智能来处理过去由人类完成的所有这些工作——eFinancialCareers，2025 年 7 月 16 日访问，https://www.efinancialcareers.com/news/jpmorgan-ai
29. JPMorgan Chase leads banking sector in AI adoption: report – CIO Dive, accessed July 16, 2025, https://www.ciodive.com/news/jpmorgan-chase-capital-one-ai-adoption-leaders-evident/730208/
    摩根大通在银行业率先采用人工智能：报道——CIO Dive，2025 年 7 月 16 日访问，https://www.ciodive.com/news/jpmorgan-chase-capital-one-ai-adoption-leaders-evident/730208/
30. JPMorgan says AI helped boost sales, add clients in market turmoil – The Economic Times, accessed July 16, 2025, https://m.economictimes.com/tech/artificial-intelligence/jpmorgan-says-ai-helped-boost-sales-add-clients-in-market-turmoil/articleshow/120906381.cms
    摩根大通表示，人工智能帮助提升了销售额，在市场动荡中增加了客户——经济时报，2025 年 7 月 16 日访问，https://m.economictimes.com/tech/artificial-intelligence/jpmorgan-says-ai-helped-boost-sales-add-clients-in-market-turmoil/articleshow/120906381.cms
31. Why JPMorgan’s AI Transformation Makes Every Other Bank Look Like They’re Operating in Slow Motion, accessed July 16, 2025, https://www.chiefaiofficer.com/post/jpmorgan-chase-ai-transformation-400-use-cases-200000-employees
    为什么摩根大通的 AI 转型让其他所有银行看起来都像是在慢动作运营，2025 年 7 月 16 日访问，https://www.chiefaiofficer.com/post/jpmorgan-chase-ai-transformation-400-use-cases-200000-employees
32. AI Boosting Payments Efficiency & Cutting Fraud | J.P. Morgan, accessed July 16, 2025, https://www.jpmorgan.com/insights/payments/payments-optimization/ai-payments-efficiency-fraud-reduction
    AI 提升支付效率与减少欺诈 | 摩根大通，访问于 2025 年 7 月 16 日，https://www.jpmorgan.com/insights/payments/payments-optimization/ai-payments-efficiency-fraud-reduction
33. Why JPMorgan Chase is Aggressively Hiring in AI – AIM Media House, accessed July 16, 2025, https://aimresearch.co/market-industry/why-jpmorgan-chase-is-aggressively-hiring-in-ai
    为何摩根大通积极招聘 AI 人才 – AIM 媒体集团，访问于 2025 年 7 月 16 日，https://aimresearch.co/market-industry/why-jpmorgan-chase-is-aggressively-hiring-in-ai
34. Why JPMorgan Chase is Aggressively Hiring in AI – AIM Media House, accessed July 16, 2025, https://aimmediahouse.com/market-industry/why-jpmorgan-chase-is-aggressively-hiring-in-ai
    为何摩根大通积极招聘 AI 人才 – AIM 媒体集团，访问于 2025 年 7 月 16 日，https://aimmediahouse.com/market-industry/why-jpmorgan-chase-is-aggressively-hiring-in-ai
35. JPMorgan Chase hits pause on hiring, bets big on AI efficiency – Storyboard18, accessed July 16, 2025, https://www.storyboard18.com/brand-marketing/jpmorgan-chase-hits-pause-on-hiring-bets-big-on-ai-efficiency-66475.htm
    摩根大通暂停招聘，押注 AI 效率 – Storyboard18，访问于 2025 年 7 月 16 日，https://www.storyboard18.com/brand-marketing/jpmorgan-chase-hits-pause-on-hiring-bets-big-on-ai-efficiency-66475.htm
36. JPMorgan Chase wants bosses to ‘resist’ hiring and use AI – Quartz, accessed July 16, 2025, https://qz.com/jpmorgan-chase-hiring-ai-jamie-dimon-1851781327
    摩根大通希望主管们“抵制”招聘并使用人工智能——石英，访问于 2025 年 7 月 16 日，https://qz.com/jpmorgan-chase-hiring-ai-jamie-dimon-1851781327
37. JPMorgan instruct managers to pause hiring and use AI instead – People Matters, accessed July 16, 2025, https://www.peoplematters.in/news/recruiting-and-onboarding/jpmorgan-instruct-managers-to-pause-hiring-and-use-ai-instead-45620
    摩根大通指示经理暂停招聘并使用人工智能——人才事务，访问于 2025 年 7 月 16 日，https://www.peoplematters.in/news/recruiting-and-onboarding/jpmorgan-instruct-managers-to-pause-hiring-and-use-ai-instead-45620
38. JPMorgan Chase Workforce Strategy: AI, Compliance & Reskilling – JobsPikr, accessed July 16, 2025, https://www.jobspikr.com/blog/jpmorgan-global-compliance-workforce-strategy/
    摩根大通劳动力战略：人工智能、合规与再培训——工作 pikr，访问于 2025 年 7 月 16 日，https://www.jobspikr.com/blog/jpmorgan-global-compliance-workforce-strategy/
39. AI’s Impact on Jobs: Reskill, Redeploy, and Invest – AInvest, accessed July 16, 2025, https://www.ainvest.com/news/ai-s-impact-on-jobs-reskill-redeploy-and-invest-250210100936aa566c7ea015/
    人工智能对工作的影响：再培训、重新部署和投资——AInvest，访问于 2025 年 7 月 16 日，https://www.ainvest.com/news/ai-s-impact-on-jobs-reskill-redeploy-and-invest-250210100936aa566c7ea015/
40. ReEntry Programs | JPMorgan Chase & Co., accessed July 16, 2025, https://careers.jpmorgan.com/us/en/students/programs/reentry-program
    重返计划 | 摩根大通公司，访问于 2025 年 7 月 16 日，https://careers.jpmorgan.com/us/en/students/programs/reentry-program
41. JPMorgan Chase Tops First-Of-Its-Kind AI Ranking, accessed July 16, 2025, https://www.jpmorgan.com/technology/news/top-ai-ranking
    摩根大通位列首个 AI 排名榜首，访问于 2025 年 7 月 16 日，https://www.jpmorgan.com/technology/news/top-ai-ranking
42. Powering Growth with Curiosity and Heart Annual Report 2023 – JPMorgan Chase, accessed July 16, 2025, https://www.jpmorganchase.com/content/dam/jpmc/jpmorgan-chase-and-co/investor-relations/documents/annualreport-2023.pdf
    以好奇心与热情驱动增长 2023 年年度报告 – 摩根大通，访问于 2025 年 7 月 16 日，https://www.jpmorganchase.com/content/dam/jpmc/jpmorgan-chase-and-co/investor-relations/documents/annualreport-2023.pdf
43. Bank of America to Spend $4 Billion on AI and New Tech Initiatives in 2025 | PYMNTS.com, accessed July 16, 2025, https://www.pymnts.com/news/banking/2025/bofa-to-spend-4-billion-on-ai-and-new-tech-initiatives-in-2025/
    美国银行将在 2025 年投入 40 亿美元用于 AI 和新科技计划 | PYMNTS.com，访问于 2025 年 7 月 16 日，https://www.pymnts.com/news/banking/2025/bofa-to-spend-4-billion-on-ai-and-new-tech-initiatives-in-2025/
44. AI Adoption by BofA’s Global Workforce Improves Productivity, Client Service | Press Releases | Newsroom | Bank of America, accessed July 16, 2025, https://newsroom.bankofamerica.com/content/newsroom/press-releases/2025/04/ai-adoption-by-bofa-s-global-workforce-improves-productivity–cl.html
    花旗集团全球员工采用 AI 提升生产力、客户服务 | 新闻稿 | 新闻中心 | 花旗集团，访问于 2025 年 7 月 16 日，https://newsroom.bankofamerica.com/content/newsroom/press-releases/2025/04/ai-adoption-by-bofa-s-global-workforce-improves-productivity–cl.html
45. A Peek At Bank Of America’s AI Playbook – Forrester, accessed July 16, 2025, https://www.forrester.com/blogs/a-peek-at-bank-of-americas-ai-playbook/
    一窥花旗集团的 AI 战略 - Forrester，访问于 2025 年 7 月 16 日，https://www.forrester.com/blogs/a-peek-at-bank-of-americas-ai-playbook/
46. AI Patents at BofA Increase 94% Since 2022 | Press Releases – Bank of America newsroom, accessed July 16, 2025, https://newsroom.bankofamerica.com/content/newsroom/press-releases/2024/10/ai-patents-at-bofa-increase-94–since-2022.html
    花旗集团自 2022 年以来 AI 专利增长 94% | 新闻稿 - 花旗集团新闻中心，访问于 2025 年 7 月 16 日，https://newsroom.bankofamerica.com/content/newsroom/press-releases/2024/10/ai-patents-at-bofa-increase-94–since-2022.html
47. Goldman Sachs Rolls Out AI Assistant Firmwide to Boost Employee Productivity, accessed July 16, 2025, https://odsc.medium.com/goldman-sachs-rolls-out-ai-assistant-firmwide-to-boost-employee-productivity-73124a7ac935
    高盛全面推行 AI 助手以提升员工生产力，访问于 2025 年 7 月 16 日，https://odsc.medium.com/goldman-sachs-rolls-out-ai-assistant-firmwide-to-boost-employee-productivity-73124a7ac935
48. Inside Goldman Sachs’ Big Bet on AI at Scale – PYMNTS.com, accessed July 16, 2025, https://www.pymnts.com/artificial-intelligence-2/2025/inside-goldman-sachs-big-bet-on-ai-at-scale/
    在高盛大规模押注人工智能——PYMNTS.com，访问于 2025 年 7 月 16 日，https://www.pymnts.com/artificial-intelligence-2/2025/inside-goldman-sachs-big-bet-on-ai-at-scale/
49. AI in Morgan Stanley: Reshaping the Future of Financial Services …, accessed July 16, 2025, https://ctomagazine.com/ai-in-morgan-stanley-shaping-the-future-of-financial-services/
    摩根士丹利中的 AI：重塑金融服务未来……，访问于 2025 年 7 月 16 日，https://ctomagazine.com/ai-in-morgan-stanley-shaping-the-future-of-financial-services/
50. Morgan Stanley Research Announces AskResearchGPT, accessed July 16, 2025, https://www.morganstanley.com/press-releases/morgan-stanley-research-announces-askresearchgpt
    摩根士丹利研究宣布推出 AskResearchGPT，访问于 2025 年 7 月 16 日，https://www.morganstanley.com/press-releases/morgan-stanley-research-announces-askresearchgpt
51. Morgan Stanley uses AI evals to shape the future of financial services – OpenAI, accessed July 16, 2025, https://openai.com/index/morgan-stanley/
    摩根士丹利利用 AI 评估塑造金融服务未来——OpenAI，访问于 2025 年 7 月 16 日，https://openai.com/index/morgan-stanley/
52. Goldman Sachs Adds AI Developer Devin to Engineering Team – eWEEK, accessed July 16, 2025, https://www.eweek.com/news/goldman-sachs-ai-coding-agent-devin/
    高盛聘请 AI 开发者 Devin 加入工程团队 – eWEEK，访问于 2025 年 7 月 16 日，https://www.eweek.com/news/goldman-sachs-ai-coding-agent-devin/
53. GOLDMAN SACHS’ AI REVOLUTION: Autonomous Coder Replacing Human Developers | Wall Street’s AI Future, accessed July 16, 2025, https://www.youtube.com/watch?v=JT80B6kKy_o
    高盛的 AI 革命：自主编码器取代人类开发者 | 华尔街的 AI 未来，访问于 2025 年 7 月 16 日，https://www.youtube.com/watch?v=JT80B6kKy_o
54. Goldman Sachs Invests in Generative AI Tools – Markets Media, accessed July 16, 2025, https://www.marketsmedia.com/goldman-sachs-invests-in-generative-ai-tools/
    高盛投资生成式 AI 工具 – Markets Media，访问于 2025 年 7 月 16 日，https://www.marketsmedia.com/goldman-sachs-invests-in-generative-ai-tools/
55. Morgan Stanley – Digital Transformation Strategies – GlobalData, accessed July 16, 2025, https://www.globaldata.com/store/report/morgan-stanley-enterprise-tech-analysis/
    摩根士丹利 – 数字化转型策略 – GlobalData，访问于 2025 年 7 月 16 日，https://www.globaldata.com/store/report/morgan-stanley-enterprise-tech-analysis/
56. Bank of America plans to spend $4 billion on AI and other new tech this year – Quartz, accessed July 16, 2025, https://qz.com/bank-of-america-bofa-ai-tech-spending-4-billion-q2-2024-1851594216
    美国银行计划今年在人工智能和其他新技术上投入 40 亿美元——石英，访问于 2025 年 7 月 16 日，https://qz.com/bank-of-america-bofa-ai-tech-spending-4-billion-q2-2024-1851594216
57. Artificial Intelligence Research – J.P. Morgan, accessed July 16, 2025, https://www.jpmorgan.com/technology/artificial-intelligence
    人工智能研究——摩根大通，访问于 2025 年 7 月 16 日，https://www.jpmorgan.com/technology/artificial-intelligence
58. AI and Model Risk Governance – J.P. Morgan, accessed July 16, 2025, https://www.jpmorgan.com/technology/news/ai-and-model-risk-governance
    人工智能与模型风险治理——摩根大通，访问于 2025 年 7 月 16 日，https://www.jpmorgan.com/technology/news/ai-and-model-risk-governance
59. Infrastructure for Trust: JP Morgan’s Open Letter to Third-Party Suppliers Signaling a New Standard for Technology Procurement – techUK, accessed July 16, 2025, https://www.techuk.org/resource/infrastructure-for-trust-jp-morgan-s-open-letter-to-third-party-suppliers-signaling-a-new-standard-for-technology-procurement.html
    信任的基础设施：摩根大通致第三方供应商的公开信标志着技术采购的新标准——techUK，访问于 2025 年 7 月 16 日，https://www.techuk.org/resource/infrastructure-for-trust-jp-morgan-s-open-letter-to-third-party-suppliers-signaling-a-new-standard-for-technology-procurement.html
60. AI’s Speed Trap: What JPMorgan Chase’s Open Letter Means for …, accessed July 16, 2025, https://ai.northeastern.edu/news/ais-speed-trap-what-jpmorgan-chases-open-letter-means-for-every-business
    AI 的速度陷阱：摩根大通公开信对……的意义，2025 年 7 月 16 日访问，https://ai.northeastern.edu/news/ais-speed-trap-what-jpmorgan-chases-open-letter-means-for-every-business
61. JPMorgan rolls out Gen AI research analyst to employees – TechInformed, accessed July 16, 2025, https://techinformed.com/jpmorgan-rolls-out-gen-ai-research-analyst-to-employees/
    摩根大通向员工推出生成式 AI 研究分析师 – TechInformed，2025 年 7 月 16 日访问，https://techinformed.com/jpmorgan-rolls-out-gen-ai-research-analyst-to-employees/
62. JPMorgan Chase’s Gen AI implementation: 450 use cases and lessons learned – Tearsheet, accessed July 16, 2025, https://tearsheet.co/artificial-intelligence/jpmorgan-chases-gen-ai-implementation-450-use-cases-and-lessons-learned/
    摩根大通生成式 AI 实施：450 个用例和经验教训 – Tearsheet，2025 年 7 月 16 日访问，https://tearsheet.co/artificial-intelligence/jpmorgan-chases-gen-ai-implementation-450-use-cases-and-lessons-learned/
63. Applied Research – J.P. Morgan, accessed July 16, 2025, https://www.jpmorgan.com/technology/applied-research
    应用研究 – 摩根大通，2025 年 7 月 16 日访问，https://www.jpmorgan.com/technology/applied-research
64. JPMorgan Chase, Toshiba and Ciena Build the First Quantum Key Distribution Network Used to Secure Mission-Critical Blockchain Application, accessed July 16, 2025, https://www.jpmorganchase.com/newsroom/press-releases/2022/jpmc-toshiba-ciena-build-first-quantum-key-distribution-network
    摩根大通、东芝和 Ciena 共同构建首个量子密钥分发网络，用于保障关键区块链应用安全，数据访问时间：2025 年 7 月 16 日，https://www.jpmorganchase.com/newsroom/press-releases/2022/jpmc-toshiba-ciena-build-first-quantum-key-distribution-network
65. JPMorgan Just Beat Big Tech to a Quantum Breakthrough – Observer, accessed July 16, 2025, https://observer.com/2025/03/jpmorgan-randomness-quantum-computing-breakthrough/
    摩根大通在量子技术领域领先科技巨头——观察者，数据访问时间：2025 年 7 月 16 日，https://observer.com/2025/03/jpmorgan-randomness-quantum-computing-breakthrough/
66. JPMorganChase, Quantinuum, Argonne National Laboratory, Oak Ridge National Laboratory and University of Texas at Austin advance the application of quantum computing to potential real-world use cases beyond the capabilities of classical computing, accessed July 16, 2025, https://www.jpmorgan.com/technology/news/certified-randomness
    摩根大通、Quantinuum、阿贡国家实验室、橡树岭国家实验室和德克萨斯大学奥斯汀分校推动量子计算在超越经典计算能力的潜在实际应用中的应用，数据访问日期为 2025 年 7 月 16 日，https://www.jpmorgan.com/technology/news/certified-randomness
67. Quantinuum, JPMorgan Chase and partners hail quantum computing breakthrough, accessed July 16, 2025, https://www.businessweekly.co.uk/posts/quantinuum-jpmorgan-chase-and-partners-hail-quantum-computing-breakthrough
    Quantinuum、摩根大通及其合作伙伴欢呼量子计算突破，访问日期：2025 年 7 月 16 日，https://www.businessweekly.co.uk/posts/quantinuum-jpmorgan-chase-and-partners-hail-quantum-computing-breakthrough
68. How JPMorgan is harnessing generative AI for customer satisfaction – TechGig CIO, accessed July 16, 2025, https://cio.techgig.com/technology/how-jpmorgan-is-harnessing-generative-ai-for-customer-satisfaction/articleshow_b2b/112424051.cms
    摩根大通如何利用生成式 AI 提升客户满意度——TechGig CIO，访问日期：2025 年 7 月 16 日，https://cio.techgig.com/technology/how-jpmorgan-is-harnessing-generative-ai-for-customer-satisfaction/articleshow_b2b/112424051.cms
69. JPMorganChase: Leadership in the Age of GenAI – Case – Faculty & Research, accessed July 16, 2025, https://www.hbs.edu/faculty/Pages/item.aspx?num=67230
    摩根大通：在生成式 AI 时代的领导力——案例研究——学院与研究，访问日期：2025 年 7 月 16 日，https://www.hbs.edu/faculty/Pages/item.aspx?num=67230
70. JPMorgan Preparing to Charge FinTechs for Consumer Bank Data | PYMNTS.com, accessed July 16, 2025, https://www.pymnts.com/news/banking/2025/jpmorgan-preparing-to-charge-fintechs-for-consumer-bank-data/
    摩根大通准备向金融科技公司收取消费者银行数据费用 | PYMNTS.com，访问日期：2025 年 7 月 16 日，https://www.pymnts.com/news/banking/2025/jpmorgan-preparing-to-charge-fintechs-for-consumer-bank-data/
71. JPMorgan to charge fintechs for customer data access – Retail Banker International, accessed July 16, 2025, https://www.retailbankerinternational.com/news/jpmorgan-fintechs-customer-data-access/
    摩根大通将向金融科技企业收取客户数据访问费用——零售银行家国际，2025 年 7 月 16 日访问，https://www.retailbankerinternational.com/news/jpmorgan-fintechs-customer-data-access/