What proprietary AI has JPMorgan Chase built in-house?
JPMorgan Chase builds the platform layer and most of its AI applications itself while renting frontier models. Its flagship is LLM Suite, which the bank calls its 'model-agnostic generative AI platform': launched in 2024 to more than 200,000 employees, it gives staff models from OpenAI, Anthropic and others in a controlled environment, plus shared capabilities for building use cases. Global CIO Lori Beer said in October 2026 that more than 200 applications run on it. On top sit named in-house tools. In asset and wealth management the 2025 annual report, published April 2026, describes SpectrumIQ, a proprietary suite of AI capabilities covering about 90,000 securities and 22 million documents; Connect Coach, with 12,000 users and 25 specialized AI agents; Smart Monitor, which cuts time from research to insight by 80%; and Proxy Voting powered by SpectrumIQ, which fully replaced external proxy advisors for US votes as of April 1, 2026. Coach AI lets private-bank advisers find information up to 95% faster (May 2025), and J.P. Morgan Payments built Turbo, an internally developed LLM tool that turns natural-language requirements into payment test scenarios. AI Research, led by Sumitra Ganesh since February 2026, works on foundation models for the financial domain. The bank has not said it trains its own large language model; Jamie Dimon wrote in April 2026 that no single AI tool will dominate.[1][2][3][4][5][6][7][8][9][10]
Data as of Oct 22, 2026 · JPMorgan profile updated Oct 21, 2026 · 32 sources cited on this page · 140 banks profiled on this site · Markdown
Which AI platforms, models and tools has JPMorgan Chase built in-house?
Systems the bank or reporters describe as JPMorgan's own, with the models they name. Figures are as of the date given; the asset and wealth management rows come from Mary Erdoes's letter in the 2025 annual report.
| Platform or model | What it does | Built on | Since | Status |
|---|---|---|---|---|
| LLM Suite | Employee portal and abstraction layer: generative AI on the desktop in a controlled environment that protects customer and company data, plus shared capabilities businesses use to build their own use cases; more than 65,000 CIB colleagues actively use it | Third-party models from OpenAI, Anthropic and others, swapped in and out | 2024 | In production; more than 230,000 users per Forbes (July 2026)[2][3][5][11] |
| SpectrumIQ | Proprietary suite of AI capabilities embedded in the Spectrum platform, tying together research, data and risk; data coverage expanded from 8,000 to about 90,000 securities and 22 million documents; about 7,000 broker research reports ingested a day | J.P. Morgan Asset Management's Spectrum platform | Described April 2026 | In production[5] |
| Proxy Voting powered by SpectrumIQ | Stewardship workflow that aggregates and analyzes proprietary data from more than 3,000 company meetings a year in US equity markets; the bank says it is the first major asset manager to fully disengage from external proxy advisors in US voting | SpectrumIQ | Q1 2026; full implementation April 1, 2026 | In production[5] |
| Connect Coach | Pushes personalized outreach ideas to advisors and supports meeting preparation, portfolio analysis and call summaries; 12,000 users across the Private Bank and US Wealth Management; 1 million AI-driven insights to 5,000 GPB users | In-house platform with 25 specialized AI agents | Launched just over a year before April 2026 | In production[5] |
| Smart Monitor | Investor AI assistant that learns investor preferences and pushes prioritized, explainable insights in real time; 80% less time from manual research to insight | In-house | Described April 2026 | In production[5] |
| Coach AI | Private-client advisers locate content and research for client conversations up to 95% faster; the bank expects advisers to expand client rosters by 50% in three to five years | In-house | Reported May 5, 2025 | In production[6] |
| Turbo | Automates payment system testing across products: turns natural-language requirements into test scenarios, with requirement parsing, automated test-data generation and retrieval-augmented context; won three Digital Banker Global AI Innovation Awards | Large language models; built by the Testing Innovation and Intelligence organization in Payments Technology | Not dated on the page | In use[7] |
| COiN, EVEE and an in-house coding assistant | COiN automates analysis of legal documents; EVEE is an intelligent call-center assistant; engineers have a proprietary coding assistant tuned for tasks such as migrating legacy systems | In-house, as reported by Forbes | Reported July 2026 | In use[11] |
| IndexGPT | Thematic investment baskets: one model generates keywords for a theme and a separate language model scans news to pick companies | OpenAI's GPT-4 | May 2024 | Launched[12] |
| Investing agents | AI agents that beat a 60/40 portfolio in backtests | JPMorgan research | July 2026 | Research result, not a product[13] |
Which JPMorgan Chase teams build its proprietary AI?
Central teams and business-line builders as named by the bank and the press. An AI-specific headcount is not published.
| Team | What it builds | Named leaders | Source date |
|---|---|---|---|
| Global Technology | More than 65,000 technology employees; Beer says many of the bank's current agents are built by her own engineering teams; more than 200 applications on LLM Suite | Lori Beer, Global CIO | October 1, 2026[4] |
| Chief Data & Analytics Office | Built and runs LLM Suite and the firm's AI program | Teresa Heitsenrether (retiring end-2026); CTO Scot Baldry taking the title; Derek Waldron, Chief Analytics Officer | June 2025-July 2026[14][15][8] |
| AI Research | AI agents and hybrid reasoning, foundation models for the financial domain, multimodal document processing, synthetic data, AI trust and safety | Sumitra Ganesh, head since February 2026 | Retrieved October 2026[9][8] |
| Asset & Wealth Management technology | SpectrumIQ, Connect Coach, Smart Monitor, Coach AI and proxy-voting workflow | Mary Erdoes, CEO; Mike Urciuoli, CIO | May 2025-April 2026[5][6] |
| Payments Technology, Testing Innovation and Intelligence | Turbo, the LLM-based payment testing platform | Not named on the page | February 2026[7] |
| Commercial & Investment Bank | Transaction screening that reviews more than double the volume with half the manual operator checks; cash-flow forecasting for treasury clients; over 90% of engineers on AI code assistants | CIB letter, 2025 annual report | April 2026[5] |
What does JPMorgan Chase say about building versus buying AI?
The bank builds the layer and buys the models. The 2025 investor day described LLM Suite as a 'flagship model-agnostic generative AI platform', and the COO's letter says it lets businesses integrate generative AI 'by leveraging a shared set of capabilities available on the platform'. In his April 2026 letter Dimon wrote that there will be a wide variety of AI models, open and closed, large and small, and that 'no single tool will dominate'.[1][2][10]
Applications are built inside the businesses. The CIB letter says that rather than a one-size-fits-all approach, 'each of our businesses has its own AI strategy aligned to the end-to-end client journey', and the asset and wealth management letter calls SpectrumIQ 'our proprietary suite of AI capabilities'. Where an outside model does the core work it is named: IndexGPT uses OpenAI's GPT-4.[5][12]
The bank keeps outside agents at arm's length. On the October 13, 2026 call Dimon said JPMorgan is 'not against agents or for agents' and warned about 'rogue agents', while its own agents, Connect Coach's 25 among them, run inside the bank's controls.[16][17][5]
What has JPMorgan Chase not disclosed about its proprietary AI?
Which models are approved for which LLM Suite tasks, the split of traffic between providers, and whether AI Research's foundation-model work has produced a model in production are not published. No patent we found names LLM Suite, SpectrumIQ or Connect Coach; the bank's LLM and agent filings are listed on the research and patents page.[3][9][18]
Usage figures for COiN and EVEE are not given in the Forbes account, Turbo's launch date and test-cycle savings are not stated, and the bank reports AI value in aggregate rather than per system.[11][7]
What does the JPMorgan profile record on this topic?
The rows below are the entries of the JPMorgan AI strategy profile that bear on this page, as of Oct 21, 2026; the profile holds the full record of 46 dated moves, 16 use cases and 73 sources.
| Use case | What the record shows | Status | Source |
|---|---|---|---|
| LLM Suite and agentic workflows | Firm-wide gen-AI platform; agents in production for software testing, private-banking research screening and payments; long-running agents planned for 2026. | In production | [19][20][21] |
| Call-centre and client assistants | Generative AI in call centres for agents serving Chase customers; AI concierge for clients is the stated end state. | Rolling out | [22][23] |
| AI investing agents (research) | Research built multi-agent systems that beat a 60/40 portfolio in two-decade backtests — a research result, not a product. | Pilot | [13] |
| Blocking outside AI agents that use customer credentials | The bank does not let third parties log in with customers' credentials, so outside agents can at most read data through open-banking connectors and cannot move money; no timeline for direct agent access. | In production | [24][25] |
| Advisor tools: Connect Coach and Smart Monitor | Connect Coach suggests next-best actions to advisors; it began with 4,600 private-banking advisors, was extended to 7,600 Chase Wealth Management staff in May 2025 and was heading for about 17,000 users by end-2025; Smart Monitor summarises filings and market moves for analysts and portfolio managers. | Rolling out | [1] |
| Consumer travel agent | Lake says the bank aims to have a consumer-facing travel agent in pilot before the end of 2026, while expecting customers not to delegate purchasing to agents soon. | Announced | [26] |
| Date | Move | Detail | Source |
|---|---|---|---|
| Oct 13, 2026 | Q3 2026 call: outside AI agents stay blocked; AI spend put near $700 million | Dimon says the bank will not let outside agents operate inside accounts and warns of a 'liability transfer' to customers who use them; the Smart Cash pilot is under way; Barnum puts annual AI spend nearer $700 million than the $2 billion cited a year earlier, calls that figure stale and says AI benefits are not yet visible in results. | [24][17][25] |
| Oct 6, 2026 | Evident AI Index: first for the fourth consecutive year | Evident's co-CEO attributes the lead to enterprise-wide strategy reaching all 22 lines of business and 10 functions, and says leading banks are building control 'harnesses' for agents. | [27] |
| Sep 28, 2026 | Joins Nvidia-led push for open-source AI agent safety tools | JPMorgan signs on to a shared, open-source safety effort for AI agents led by Nvidia, as the industry works out how to give agents more autonomy without losing control. | [28] |
| Jul 2, 2026 | AI chief Heitsenrether to retire | The chief data and analytics officer, who led the LLM Suite build-out, leaves at year-end; her title and portfolio go to CTO Scot Baldry. | [15] |
| Jun 9, 2026 | Long-running agents promised for 2026 | Waldron: agents will run for an hour or two, act as 'team managers' delegating to sub-agents; AI already lifted private-banking gross sales 20%. | [19] |
| Mar 23, 2026 | Engineering post on securing agents | The bank publishes its approach: align safeguards to capability and risk; agents combining untrusted inputs, sensitive data and authority to act require continuous enforcement. | [29] |
| Feb 24, 2026 | Company update: 150,000 weekly LLM Suite users, use cases doubled | Dimon says 150,000 people use the platform every week and think they save about four hours a week; Barnum says AI use cases in production doubled and about $600 million of identified efficiencies are partly AI-related; Mary Erdoes says one controls-review AI tool spread from 200 to 3,000 staff. | [30][31] |
| Jun 3, 2025 | LLM Suite named American Banker Innovation of the Year | The bank's post says the platform went from zero to 200,000 onboarded users in eight months after its summer 2024 release; Derek Waldron calls it the firm's planned 'AI hub for employees'. | [14] |
| May 19, 2025 | Investor Day: 200,000 LLM Suite users, 175 AI use cases in CIB | The investor day transcript reports over 200,000 employees with LLM Suite access, more than 175 AI use cases in production in the commercial and investment bank, a 40% fall in KYC unit cost since 2022, and a plan to cut consumer operations headcount about 10% over five years while the business grows more than 25%. | [1] |
| Aug 9, 2024 | LLM Suite reaches 60,000 employees | CNBC reports the opt-in rollout; by mid-2025 American Banker describes it as available to almost all employees. | [3] |
| May 3, 2024 | IndexGPT thematic baskets launch | Built with OpenAI's GPT-4: the model generates keywords for a theme and a separate language model scans news to pick companies. Rui Fernandes, head of markets trading structuring, calls it a first step in a gradual, long-term AI plan for the index line. The name had been filed as a trademark in May 2023. | [12][32] |
| Jan 1, 2023 | Employees get OpenAI models through an early LLM Suite | The groundwork for the firm-wide platform: a corporate ChatGPT for drafting emails and summarising documents. | [23] |
Does JPMorgan Chase build its own AI models?
It builds platforms and applications rather than a public foundation model. LLM Suite, its flagship generative AI platform, is model-agnostic and serves models from OpenAI, Anthropic and others; IndexGPT uses OpenAI's GPT-4. Its AI Research group works on foundation models for the financial domain, and tools such as SpectrumIQ, Connect Coach and Turbo are built in-house, but the bank has not announced a large language model of its own.
What is JPMorgan's LLM Suite?
LLM Suite is JPMorgan Chase's proprietary generative AI platform, launched in 2024 to more than 200,000 employees. It gives staff generative AI on the desktop in a controlled environment that protects customer and company data and offers shared capabilities so businesses can build their own use cases. Lori Beer said in October 2026 that more than 200 applications are built on it.
What are SpectrumIQ and Connect Coach?
Both are J.P. Morgan Asset & Wealth Management tools described in the 2025 annual report. SpectrumIQ is a proprietary suite of AI capabilities inside the Spectrum platform that ties together research, data and risk across about 90,000 securities and 22 million documents, and now powers the firm's proxy voting. Connect Coach serves 12,000 advisors and bankers with 25 specialized AI agents that push personalized outreach ideas and summarize calls.
Which outside AI models does JPMorgan use?
LLM Suite carries models from OpenAI, Anthropic and other providers, swapped in and out under the bank's controls, and IndexGPT was built with OpenAI's GPT-4. Dimon wrote in April 2026 that there will be many AI models, open and closed, and that no single tool will dominate, which is the bank's reason for staying model-agnostic.
Who builds JPMorgan Chase's AI?
Global CIO Lori Beer's technology organization of more than 65,000 people builds many of the agents; the chief data and analytics office, led by Teresa Heitsenrether until her retirement at the end of 2026 and then by CTO Scot Baldry, built LLM Suite; AI Research is led by Sumitra Ganesh. Business lines build their own tools, such as Asset & Wealth Management's Connect Coach and Payments Technology's Turbo.
- 2025 Investor Day Transcript - Full Transcript
- Future-Proofing the Company and Our Operations (COO letter, 2024 annual report)
- How JPMorganChase democratized employee access to gen AI
- The Most Powerful Women in Banking, #5, Lori Beer, JPMorganChase
- Line of Business CEO Letters to Shareholders, Annual Report 2025
- JPMorgan says AI helped boost sales, add clients in market turmoil
- Turbo Wins 3 Categories in Global AI Innovation Awards
- JPMorgan's AI leadership is getting a reset. These are the key players.
- Artificial Intelligence Research
- Jamie Dimon's Letter to Shareholders, Annual Report 2025
- How JPMorgan Chase Is Building The AI-Powered Bank Of The Future
- JPMorgan unveils IndexGPT in next Wall Street bid to tap AI boom
- JPMorgan Builds AI Agents That Beat 60/40 Portfolio in Backtests
- LLM Suite named 2025 'Innovation of the Year' by American Banker
- JPMorgan AI chief Teresa Heitsenrether retiring in 2026
- Jamie Dimon Says JPMorgan Isn't Using External AI Agents
- JPMorgan CEO Dimon warns customers of 'liability transfer' with AI
- US20260050765A1 - Method and system for an AI agent framework performing workflow analytics
- JPMorgan Chase plans to deploy more powerful AI agents this year
- Translating technology from cost to competitive edge (Gill Haus)
- JPMorgan Payments' agentic commerce play
- The Rise of Artificial Intelligence at JPMorgan
- Here's JPMorgan Chase's blueprint to become the world's first fully AI-powered megabank
- JPMorgan Chase blocks outside AI agents like Muse, Instinct from directly accessing accounts
- JPMorgan warns of risks to consumers who rely on AI agents
- JPMorgan's Lake sees longer road to agentic commerce adoption
- JPMorganChase still tops AI ranking as big-bank lead widens
- JPMorgan Stock Slips 1% as Bank Agents Get an Open Safety Layer
- Securing the next generation of AI agents
- 2026 Company Update Transcript - Chairman and CEO Q&A
- 2026 Company Update Transcript (full event)
- JPMorgan taps ChatGPT for new thematic investment suite
When JPMorgan moves on AI, you'll read it here first.
when one of the banks we cover moves on AI, it is in the next morning's brief · six sourced stories · 7 am PT · free
plus every regulation, bank and agent page update, the morning after · leave any morning