How does Goldman Sachs govern AI and manage its risks?
Goldman Sachs governs AI through its risk committees and board rather than a published set of AI principles. The 2026 proxy statement says the risks of 'generative and agentic artificial intelligence' were 'top of mind' for the board and its committees in 2025, and that the firm's strategy is 'empowered by the responsible deployment of technology, such as artificial intelligence'. The 2025 10-K, filed February 2026, carries a standalone AI risk factor covering generative-AI errors, bias, opacity, reliance on third-party models and heightened risk from OneGS 3.0. Oversight has moved up: the Q1 2026 10-Q placed an AI Risk and Controls Committee under the Firmwide Technology Risk Committee, and by the Q2 10-Q it had become a Firmwide AI Risk Committee co-chaired by the CIO and the chief risk officer. In practice the GS AI Platform applies prompt controls that flag sensitive inputs. On October 13, 2026, CEO David Solomon called for 'rigorous testing, independent review, continuous monitoring and clear human accountability', saying the industry's experience of testing and independent review offers a framework for deploying AI at scale. Partner Chris Churchman has warned of 'cognitive atrophy' if bankers outsource reasoning to models.[1][2][3][4][5][6][7][8][9]
Data as of Oct 16, 2026 · Goldman profile updated Oct 13, 2026 · 14 sources cited on this page · 140 banks profiled on this site · Markdown
What governance structures does Goldman Sachs disclose for AI?
| Mechanism | What it says or does | Source | Date |
|---|---|---|---|
| Board oversight | Risks of new technologies 'including generative and agentic artificial intelligence, were top of mind for the Board and its Committees' | 2026 proxy statement | March 20, 2026[1] |
| Strategy framing | Investment 'empowered by the responsible deployment of technology, such as artificial intelligence' | 2026 proxy statement | March 20, 2026[1] |
| 10-K risk factor | Standalone AI risk factor: generative-AI errors, bias, opacity, third-party models, OneGS 3.0 execution | 2025 Form 10-K | February 25, 2026[2] |
| AI Risk and Controls Committee | Oversees risks from the use of AI; reports to the Firmwide Technology Risk Committee | Q1 2026 10-Q | May 1, 2026[3] |
| Firmwide AI Risk Committee | Co-chaired by the CIO and the chief risk officer | Q2 2026 10-Q | August 3, 2026[4] |
| Prompt controls | The GS AI Platform checks what staff enter in prompts and flags it | American Banker | July 25, 2025[5] |
| Industry standards | Supporter of the FINOS common-controls project for AI services | Banking Dive | June 24, 2025[10] |
What has Goldman Sachs' leadership said about AI safety and risk?
On the Q3 2026 call Solomon said aspects of AI such as 'the direction of travel, how the technology will be evolved, how it will be deployed' are 'unknown[s] we have to live with', and that 'we must enable innovation and also demand responsibility'. He said the financial industry's experience with testing and independent review provides a framework for deploying AI responsibly at scale.[6][7]
Chris Churchman, head of Marquee, warned in August 2026: 'There's a huge danger here that in the era of AI, we outsource our reasoning to these models, and we have cognitive atrophy.' In April 2026 Solomon said the firm had Anthropic's Mythos and was working with Anthropic and its security vendors to 'harness frontier capabilities', after Treasury and the Fed convened bank leaders on AI-enabled cyber risk.[9][11][12]
Goldman hired Evan Kotsovinos from Google in July 2026, where he led AI safety and security work within Privacy, Safety and Security, as head of Asset & Wealth Management engineering. Not on the record: a published AI principles document, a named chief AI ethics or responsible-AI officer, or external audits of the firm's AI controls.[13][1]
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 |
|---|---|---|---|
| Mythos vulnerability discovery | Using the restricted model and working with Anthropic on cyber risks. | In production | [12] |
| AI risk factor and Firmwide AI Risk Committee | 10-K risk factor on generative-AI output, bias, opacity and third-party models; oversight by a committee co-chaired by the CIO and chief risk officer. | In production | [2][4] |
| Date | Move | Detail | Source |
|---|---|---|---|
| Oct 13, 2026 | Q3 2026 call: Solomon sets out AI guardrails and compute-supply risk | He cites a $10 trillion US data-centre build-out through 2032, says compute demand exceeds supply, and calls for testing, independent review, monitoring and human accountability; efficiency ratio is 59.4% year to date. | [14][8] |
| Aug 24, 2026 | Partner warns of 'cognitive atrophy' | Chris Churchman, leading a flagship AI project, on the danger of outsourcing reasoning to models. | [9] |
| Aug 3, 2026 | AI risk oversight moves up to a firmwide committee | The Q2 10-Q describes a Firmwide AI Risk Committee co-chaired by the CIO and chief risk officer; the 10-K had an AI risk and controls committee under technology risk. | [4][2] |
| May 1, 2026 | Q1 10-Q: AI Risk and Controls Committee sits under Technology Risk | The committee, which oversees risks from the use of AI, reports to the Firmwide Technology Risk Committee (co-chaired by the CISO and CTO); by the Q2 10-Q it had become a Firmwide AI Risk Committee. | [3][4] |
| Apr 13, 2026 | Using Mythos; regulators call in bank chiefs | Solomon says the firm is supplementing cyber resilience after Bessent and Powell convened Wall Street leaders. | [12] |
| Feb 25, 2026 | 2025 10-K carries a standalone AI risk factor | Generative-AI errors, bias, opacity, reliance on third-party models and heightened risk from OneGS 3.0. | [2] |
Does Goldman Sachs have an AI risk committee?
Yes. The Q1 2026 10-Q described an AI Risk and Controls Committee reporting to the Firmwide Technology Risk Committee; by the Q2 2026 10-Q it had become a Firmwide AI Risk Committee co-chaired by the chief information officer and the chief risk officer.
What does Goldman Sachs' 10-K say about AI?
The 2025 Form 10-K, filed in February 2026, includes a standalone risk factor on artificial intelligence. It covers errors and bias in generative-AI output, the opacity of models, reliance on third-party model providers and the heightened risks that come with the AI-driven OneGS 3.0 operating model.
Does Goldman Sachs' board oversee AI?
The 2026 proxy statement says the risks of deploying generative and agentic AI were top of mind for the board and its committees in 2025, and that the board followed OneGS 3.0, which it describes as a new operating model propelled by AI. It does not name a board committee dedicated to AI.
What did David Solomon say about AI guardrails?
On October 13, 2026 Solomon said 'the choice is not between halting progress and moving forward without guardrails' and that the industry must enable innovation while demanding responsibility, through rigorous testing, independent review, continuous monitoring and clear human accountability.
- 2026 Proxy Statement for the Annual Meeting of Shareholders
- The Goldman Sachs Group, Inc. Annual Report on Form 10-K for 2025
- The Goldman Sachs Group, Inc. Quarterly Report on Form 10-Q for the period ended March 31, 2026
- The Goldman Sachs Group, Inc. Quarterly Report on Form 10-Q for the period ended June 30, 2026
- Goldman Sachs staff now write a million gen AI prompts a month
- What Wells Fargo, Citi and Goldman Sachs' CEOs are saying about AI
- Goldman Sachs (GS) Q3 2026 Earnings Call Transcript
- Goldman's Solomon wants 'clear human accountability' on AI
- Goldman Sachs partner warns of 'huge danger' in letting AI replace bankers' reasoning skills
- Banking industry, big tech unite to forge AI adoption guidelines
- Earnings call transcript: Goldman Sachs beats Q1 2026 forecasts, stock dips
- Goldman Is Using Mythos, Working With Anthropic on Cyber Risks
- Evan Kotsovinos Joins Goldman Sachs as Partner and Head of Asset & Wealth Management Engineering
- Earnings call transcript: Goldman Sachs tops Q3 2026 estimates as revenue rises
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