Examples
Examples from published documents
- Where marginal risk analysis "plays a major role in a decision to move forward, explicit approval of the Risk Report by the Board and LTBT (rather than just the CEO and RSO) will be required" (Anthropic RSP v3.4, section 3.4)
- "Where necessary, the Board may reverse a decision and/or mandate a revised course of action." (OpenAI Preparedness Framework v2, Appendix B)
- Raising a capability threshold "will require approval by our Board of Directors, with input from external security and AI safety advisers" (Magic AGI Readiness Policy, item 3)
Counter-examples
Not to be confused with
- Executive sign-off is not board oversight: Anthropic's CEO and Responsible Scaling Officer approve a Risk Report, and the Board and Long-Term Benefit Trust are then notified as the layer above (Anthropic RSP v3.4, section 3.4).
- A committee named 'board' is not necessarily a governing body: G42's Frontier AI Governance Board is made up of executives and proposes framework changes that its Executive Leadership Committee approves (G42 Frontier Safety Framework, section 5).
Related terms
Related AI safety terms
Terms defined under this element that are not NIKOLAI elements themselves, each as the cited document uses it. Every term is also listed in the AI safety glossary.
- Long-Term Benefit Trust
- The Long-Term Benefit Trust (LTBT) is a body named in Anthropic's Responsible Scaling Policy that receives regular briefings, approves the choice of external reviewers, may request external review of Risk Reports, and must co-approve reports that rely heavily on marginal risk analysis.
- Source: Anthropic Responsible Scaling Policy v3.4
- Safety and Security Committee
- The Safety and Security Committee (SSC) is the OpenAI board committee that, under the Preparedness Framework, has visibility into safety processes and can review decisions and require reports from OpenAI leadership; the Board may reverse a decision.
- Source: OpenAI Preparedness Framework v2
- Internal governance
- Internal governance is the set of roles, committees and review procedures a developer uses to make sure its safety framework is carried out. California SB 53 requires a frontier AI framework to describe how the developer institutes internal governance practices.
- Source: California SB 53
- Management body in its supervisory function
- The management body in its supervisory function is the EU GPAI Code of Practice's term for the board-level body, or another suitable independent body such as a council, that oversees a provider's systemic risk assessment and mitigation.
- Source: EU GPAI Code of Practice, Safety and Security Chapter
Source of record
Where this definition comes from
Anthropic Responsible Scaling Policy v3.4, §3.4, p.14
“In the event that marginal risk analysis (see previous section) plays a major role in a decision to move forward, explicit approval of the Risk Report by the Board and LTBT (rather than just the CEO and RSO) will be required.”
https://cdn.sanity.io/files/4zrzovbb/website/0bacdc8440ea96e62a8766d99ebe1d4eea6d5f3a.pdfOpenAI Preparedness Framework v2, Appendix B
“The Safety and Security Committee (SSC) of the OpenAI Board of Directors (Board) will be given visibility into processes, and can review decisions and otherwise require reports and information from OpenAI Leadership as necessary to fulfill the Board's oversight role. Where necessary, the Board may reverse a decision and/or mandate a revised course of action.”
https://cdn.openai.com/pdf/18a02b5d-6b67-4cec-ab64-68cdfbddebcd/preparedness-framework-v2.pdfEU GPAI Code of Practice, Safety and Security Chapter, Commitment 8, Measure 8.1, p.24
“Systemic risk oversight: The responsibility for overseeing the Signatory's systemic risk management processes and measures has been assigned to a specific committee of the management body in its supervisory function (e.g. a risk committee or audit committee) or one or multiple suitable independent bodies (such as councils or boards).”
https://ec.europa.eu/newsroom/dae/redirection/document/118119
Crosswalk
How named organisations use this concept
| Organisation | Their term, as published | Match & verification | Source |
|---|---|---|---|
| AnthropicShadow mapping Anthropic RSP v3.4 | “Notification: "Following approval of a Risk Report, the CEO and RSO will promptly share their decision(s), the underlying Risk Report, and internal feedback with both the Board and the LTBT." (§3.4). Approval where marginal risk analysis is central: "explicit approval of the Risk Report by the Board and LTBT (rather than just the CEO and RSO) will be required." Reviewer selection: "In selecting external reviewers, we will consult with the Board and obtain the approval of the LTBT." (§3.6.1). Briefings: "We will regularly brief the LTBT on plans and developments related to our Risk Reports" (§4(2)). Policy changes "approved by the Board in consultation with the LTBT" (§4(8)). The v3.2 changelog entry "authorizes the LTBT to request external review of Risk Reports".” Also cites the Anthropic Risk Report for exercise status: "Since this change to our RSP, the LTBT has not requested an external review (nor has the RSP required that we conduct one)" (Risk Report §1.3.5). | exactEQ confidence: high | Anthropic Responsible Scaling Policy v3.4 |
| OpenAIShadow mapping OpenAI Preparedness Framework v2 / Frontier Governance Framework | “PF App. B: the SSC "will be given visibility into processes, and can review decisions and otherwise require reports and information from OpenAI Leadership as necessary to fulfill the Board's oversight role. Where necessary, the Board may reverse a decision and/or mandate a revised course of action." FGF: "The board of directors of OpenAI Ireland Limited exercise systemic risk oversight under this Framework for EU purposes." (§6); material FGF updates "will be presented to the Safety and Security Committee of the board of directors of the OpenAI Foundation and to the board of directors of OpenAI Ireland Limited for oversight" (§7.1).” Also cites the Frontier Governance Framework and OpenAI's September 2026 post (the SSC of the Foundation board "provides governance over safety and security practices across all of OpenAI, including OpenAI Group PBC"). PF v2 (April 2025) places the SSC on 'the OpenAI Board of Directors'; the FGF and the September 2026 post place it on the OpenAI Foundation board. | exactEQ confidence: high | OpenAI Preparedness Framework v2 |
| Google DeepMindShadow mapping Frontier Safety Framework v3.1 | “"The updated version and framework assessment will be reviewed by the appropriate corporate governance bodies." (s.5.1); governance "includes legal, compliance, and safety reviews with escalation procedures to ensure appropriate oversight." (s.4.1)” DU: oversight by governance bodies is declared, but the bodies and their powers are not named. | noneDU confidence: medium | Google DeepMind Frontier Safety Framework v3.1 |
| MetaShadow mapping Meta Advanced AI Scaling Framework v2 | “"Meta's Board of Directors provides oversight of the company's product and regulatory compliance, ensuring accountability across all lines of defense." (§2.3)” BR: general board oversight of compliance; no framework-specific power (approval, review or reversal) is stated. | broadBR confidence: medium | Meta Advanced AI Scaling Framework v2 |
| MicrosoftShadow mapping Microsoft Frontier Governance Framework (Feb 2026) | “"This framework is subject to Microsoft's broader corporate governance procedures, including independent internal audit and board oversight." (s.5, p.9)” BR: the board committee involved is not named and no framework-specific power is stated. | broadBR confidence: medium | Microsoft Frontier Governance Framework (Feb 2026) |
| G42Shadow mapping G42 Frontier Safety Framework | “"A dedicated Frontier AI Governance Board, composed of our Chief Responsible AI Officer, Head of Responsible AI, Head of Technology Risk, and General Counsel, shall oversee all frontier model operations, reviewing safety protocols, risk assessments, and escalation decisions." Its duties include "escalating material incidences of non-compliance to the G42 Executive Leadership Committee" (s.5).” CL: despite the name, this is a body of executives, not a board of directors; framework changes are approved by the Executive Leadership Committee. | closeCL confidence: medium | G42 Frontier Safety Framework |
| MagicShadow mapping Magic AGI Readiness Policy | “The engineering team "is responsible for making this known immediately to the leadership team and Magic's Board of Directors (BOD)" when a threshold is crossed (item 1); "A member of staff will be appointed who is responsible for sharing the following with our Board of Directors on a quarterly basis" (item 2); threshold or benchmark changes "will require approval by our Board of Directors, with input from external security and AI safety advisers" (item 3).” The page carries the banner "Note: This policy is outdated. We're working on an update." | exactEQ confidence: medium | Magic AGI Readiness Policy |
| EUShadow mapping EU GPAI Code of Practice, Safety and Security Chapter | “Measure 8.1 responsibility "(1) Systemic risk oversight: Overseeing the Signatories' systemic risk assessment and mitigation processes and measures", allocated to "the management body in its supervisory function or another suitable independent body (such as a council or board)"; presumed fulfilled where it "has been assigned to a specific committee of the management body in its supervisory function (e.g. a risk committee or audit committee)". Assurance is provided "to the management body in its supervisory function or another suitable independent body (such as a council or board)".” | exactEQ confidence: high | EU GPAI Code of Practice, Safety and Security Chapter |
| California SB 53Shadow mapping California SB 53 | “For whistleblower disclosures only: "the disclosures and responses of the process required by this subdivision shall be shared with officers and directors of the large frontier developer at least once each quarter." (Labor Code 1107.1(e)); a director accused in a disclosure is excluded from receiving it.” NR: directors receive quarterly whistleblower reports; SB 53 otherwise leaves internal governance to the developer ('Instituting internal governance practices to ensure implementation of these processes', 22757.12(a)(9)). | narrowNR confidence: medium | California SB 53 |
| Seoul Summit signatoriesShadow mapping Frontier AI Safety Commitments (Seoul 2024) | “Commitment VI: "Adhere to the commitments outlined in I-V, including by developing and continuously reviewing internal accountability and governance frameworks and assigning roles, responsibilities and sufficient resources to do so."” BR: asks for internal accountability and governance frameworks without naming a governing body. | broadBR confidence: medium | Frontier AI Safety Commitments (AI Seoul Summit 2024) |
What do these codes mean?
- exact
- The source term is equivalent to this element
- close
- The source term is close but not equivalent to this element
- broad
- The source term is broader than this element
- narrow
- The source term is narrower than this element
- none
- No mapping claim — used for false-friend and declared-but-undefined rows
- EQ
- Equivalent
- CL
- Close
- BR
- Source is broader than the element
- NR
- Source is narrower than the element
- FF
- False friend — same or similar label, different meaning
- DU
- Declared but undefined by the source
- UV
- Unverified
Related, not mapped
Pointers that are not crosswalk claims
These sources mention this concept but do not define or map it clearly enough to count as a crosswalk row — noted here so the research is visible without overstating it as a mapping.
- Google DeepMind
Names a standing internal body, the Responsibility and Safety Council, which "works across teams to monitor emerging risk, review ethics and safety assessments and implement relevant technical and policy mitigations"; its membership and decision rights are not disclosed in the sources reviewed, and it is an internal review body rather than a board.
Google DeepMind, 'Deepening our partnership with the UK AI Security Institute'
Divergence
Where sources materially disagree
The powers differ by source: Anthropic's Board and Long-Term Benefit Trust are notified of decisions, must approve Risk Reports that rely heavily on marginal risk analysis, and (the LTBT) approve external reviewers; OpenAI's Board may reverse a decision; Magic's Board approves threshold changes; the EU Code assigns oversight to the supervisory management body or a committee of it; Meta, Microsoft and Google DeepMind describe board or governance-body oversight without framework-specific powers. G42 uses 'Board' for an executive committee.
Gap
The research did not find a board-level oversight provision in xAI's June 2026 Frontier AI Framework. Few sources say whether an oversight power has been used; Anthropic's August 2026 Risk Report is the exception found, stating that the LTBT had not requested an external review.
