What technology stack does Goldman Sachs use for AI?
Goldman Sachs runs AI through the GS AI Platform, a firm-hosted single entry point described in June 2024 as fronting Microsoft-hosted GPT-3.5 and GPT-4, Google Gemini and Meta Llama; by July 2025 staff could pick among current GPT, Gemini and Claude models, with prompt controls that flag the information users enter. The GS AI Assistant on that platform reached all 46,000 employees in July 2025. Engineering is the deepest layer: GitHub Copilot for 12,000 developers since 2024, Cognition's Devin from July 2025, and by August 2026 hundreds of agents working beside 12,000 engineers on production tasks, starting with legacy modernization; CIO Marco Argenti reported up to 20% efficiency gains. Data runs on Legend, the data-modeling platform Goldman open-sourced through FINOS in 2020, and Legend lets staff search firm files in natural language. In the cloud, Goldman built Financial Cloud for Data with AWS in 2021, and CFO Denis Coleman said in April 2026 that OneGS 3.0 is bringing forward cloud migration and data-quality work because both are critical to deploying AI. Argenti, a former AWS executive, has described how agentic platforms such as Claude Code changed AI coding inside the bank (March 2026).[1][2][3][4][5][6][7][8][9][10][11]
Data as of Oct 16, 2026 · Goldman profile updated Oct 13, 2026 · 21 sources cited on this page · 140 banks profiled on this site · Markdown
What are the main layers of Goldman Sachs' AI technology stack?
| Layer | What Goldman uses | Detail | Date |
|---|---|---|---|
| Model access | GS AI Platform | Single entry point; GPT via Microsoft, Gemini and Llama (2024); GPT, Gemini and Claude choices (2025); in-house use of ChatGPT banned | June 2024[1][2] |
| Employee assistant | GS AI Assistant | All 46,000 employees; more than a million prompts a month; summarizing, drafting and data analysis | July 2025[2][12] |
| Data controls | Prompt controls | Check what information staff enter in prompts and flag it | July 2025[2] |
| Coding assistants | GitHub Copilot | 12,000 developers | 2024[2] |
| Coding agents | Cognition Devin, Claude | Hundreds of agents beside 12,000 engineers on production work, starting with legacy modernization; developers set specs and review output | July 2025-August 2026[3][4][5] |
| Business agents | Claude agents co-built with Anthropic | Trade and transaction accounting; client vetting and onboarding | February 2026[13] |
| Data platform | Legend | Open-sourced through FINOS in October 2020; natural-language search of firm files | 2020; 2026[6][7] |
| Cloud and client data | AWS (Goldman Sachs Financial Cloud for Data) | Cloud-native data and analytics for institutional clients, extending GS Quant and PlotTool Pro | November 2021[8] |
| Client platform | Marquee AI | Natural-language assistant that breaks questions into research topics and attributes statements to sources; internal only as of August 2026 | August 2026[14] |
Does Goldman Sachs build or buy its AI?
Both. Goldman builds the platform layer itself, the GS AI Platform and Assistant, and routes to models bought from several labs. Argenti, who previously worked at Amazon Web Services, discussed the development of Goldman's own internal AI tools on Bloomberg's Odd Lots in August 2024.[1][10]
For agents it buys and co-develops: Devin from Cognition, in which Goldman holds no stake, and Claude agents built with embedded Anthropic engineers. Argenti measures idea-to-production speed rather than individual AI usage: 'There's zero time between idea and prototype.'[3][13][7]
Not disclosed: Goldman's split between cloud and its own data centers, its GPU capacity or a technology budget figure. Its patents point to in-house work on retrieval-augmented generation, embedding models and coding-assistant context.[15][16][17]
What does the Goldman profile record on this topic?
The rows below are the entries of the Goldman Sachs AI strategy profile that bear on this page, as of Oct 13, 2026; the profile holds the full record of 37 dated moves, 13 use cases and 62 sources.
| Use case | What the record shows | Status | Source |
|---|---|---|---|
| GS AI Assistant | Firm-wide conversational assistant on the GS AI Platform. | In production | [2][18] |
| Devin and coding agents | Autonomous software engineer piloted in developer teams. | Pilot | [18] |
| Coding agents on production work | Devin is broadly available to engineers and hundreds of agents work beside 12,000 developers; Argenti cites up to 20% efficiency gains and says developers now set specs and review output. | In production | [13][4][5] |
| Prompt controls on the GS AI Platform | Controls check what information staff enter in prompts and flag it; the platform lets users choose among current GPT, Gemini and Claude models. | In production | [2] |
| Frontier-lab and AI-application vendor relations | A Partner-level head of AI product management leads strategic engagement with frontier labs and AI application providers, alongside embedded Anthropic engineers and the Devin deployment. | In production | [19][13] |
| Date | Move | Detail | Source |
|---|---|---|---|
| Aug 6, 2026 | Forbes: hundreds of agents work beside 12,000 engineers | Goldman is described as the first major bank to deploy virtual software engineers on real production tasks, starting with legacy modernisation. | [4] |
| May 8, 2026 | Argenti measures idea-to-production speed, not AI usage | The CIO says tracking individual use is the wrong metric for his 12,000 engineers; Legend lets staff search firm files in natural language. | [7] |
| Apr 13, 2026 | Q1 2026 call: cloud and data spend brought forward for AI | CFO Denis Coleman says early OneGS 3.0 learnings mean accelerating cloud migration and data quality, critical to deploying AI across the firm. | [9] |
| Mar 20, 2026 | Solomon's letter flags AI-exposed private credit | 'Exposure to software companies that may be adversely affected by AI' named as a reminder that the credit cycle has not been repealed. | [20] |
| Jul 25, 2025 | GS AI Assistant reaches the whole workforce | 46,000 employees; a million prompts a month; Devin agents joining developer teams. | [2] |
| Jul 11, 2025 | Devin pilot announced: 'hundreds of Devins', maybe thousands | Argenti expects Cognition's agent to raise developer productivity three to four times over earlier AI tools; Goldman is the first major bank to use it, per Cognition, and holds no stake in Cognition. | [3] |
| Jun 23, 2025 | Argenti memo: GS AI Assistant launches firmwide | About 10,000 employees already use it; it will help with summarising documents, drafting content and data analysis. | [12] |
| Jan 21, 2025 | GS AI Assistant reaches about 10,000 employees | Argenti aims to cover all knowledge workers within the year; answers draw on OpenAI, Google Gemini and Meta Llama models, with Anthropic, Mistral and Cohere under evaluation. | [21] |
| Jun 27, 2024 | GS AI Platform: one entry point, several model vendors | Argenti says the first generative-AI tool, for code generation, reaches thousands of developers by month end; the platform uses Microsoft-hosted GPT-3.5 and GPT-4, Google Gemini and Meta Llama, and the bank had banned in-house ChatGPT. | [1] |
| Jan 1, 2024 | GitHub Copilot for 12,000 developers | The first firm-wide generative-AI deployment, to engineering. | [2] |
What is the GS AI Platform?
It is Goldman Sachs' firm-hosted gateway to generative AI. Reported in June 2024 as the single entry point for generative AI, it put Microsoft-hosted GPT-3.5 and GPT-4, Google Gemini and Meta Llama behind one interface; by July 2025 staff could choose among current GPT, Gemini and Claude models. The GS AI Assistant runs on it.
What coding tools do Goldman Sachs engineers use?
Its 12,000 developers have had GitHub Copilot since 2024. Cognition's autonomous engineer Devin joined teams in July 2025 and Forbes reported in August 2026 that hundreds of agents work beside the engineers on production tasks. CIO Marco Argenti cited up to 20% efficiency gains from the Devin pilots.
Which cloud does Goldman Sachs use?
Goldman built Goldman Sachs Financial Cloud for Data with AWS, announced at re:Invent in November 2021, for institutional clients' data and analytics. CFO Denis Coleman said in April 2026 that OneGS 3.0 is accelerating cloud migration and data-quality work because both are critical to deploying AI. The full cloud footprint is not disclosed.
What is Goldman Sachs Legend?
Legend is Goldman's data-modeling and governance platform, open-sourced through FINOS in October 2020. Inside the bank it lets staff search firm files in natural language without knowing where a document sits, according to CIO Marco Argenti in May 2026.
- Goldman Sachs Deploys Its First Generative AI Tool Across the Firm
- Goldman Sachs staff now write a million gen AI prompts a month
- Goldman Sachs is piloting its first autonomous coder in major AI milestone for Wall Street
- How Goldman Sachs Is Using Agentic AI For Software Engineering At Scale
- Marco Argenti gives Goldman Sachs coders an AI helper (Most Innovative People in Finance, #3)
- Goldman Sachs data governance code goes live on Finos
- Goldman Sachs’ tech boss says tracking individual AI usage isn’t useful
- Goldman Sachs and AWS Collaborate to Create New Data Management and Analytics Solutions for Financial Services Organizations
- Goldman Sachs (GS) Q1 2026 Earnings Transcript
- Goldman Sachs CIO on How the Bank Is Actually Using AI (Odd Lots)
- Goldman CIO Marco Argenti on the Warp-Speed Improvements in AI (Odd Lots)
- Goldman Sachs launches AI assistant firmwide, memo shows
- Goldman Sachs taps Anthropic's Claude to automate accounting, compliance roles
- Building AI Systems for Capital Markets (Goldman Sachs Exchanges)
- US12585896B1: Embedding-free retriever-augmented generation (RAG) architectures
- US12518171B2: Data generation and retraining techniques for fine-tuning of embedding models for efficient data retrieval
- US20260072655A1: Context injection for artificial intelligence (AI) coding assistants
- Marco Argenti gives Goldman Sachs coders an AI helper
- Archana Vemulapalli Joins Goldman Sachs as Partner and Head of AI Product Management and Strategic Relations
- Financial Services Roundup: Market Talk (Solomon shareholder letter)
- Goldman Sachs rolls out an AI assistant for its employees as artificial intelligence sweeps Wall Street
When Goldman 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 ET · free
plus every tracker, bank and agent page update, the morning after · leave any morning