Source of record
Where this definition comes from
California SB 53, 22757.13(i)(1)
“A frontier developer that intends to comply with this section by complying with the requirements of, or meeting the standards stated by, a federal law, regulation, or guidance document designated pursuant to subdivision (h) shall declare its intent to do so to the Office of Emergency Services.”
https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202520260SB53OpenAI Frontier Governance Framework, §1
“Under California's Transparency in Frontier AI Act (TFAIA), this FGF is our Frontier AI Framework”
https://cdn.openai.com/pdf/e37d949b-8c9f-4d76-b99e-4272f4631a7e/openai-frontier-governance-framework.pdf
Crosswalk
How named organisations use this concept
| Organisation | Their term, as published | Match | Source |
|---|---|---|---|
| OpenAI OpenAI Frontier Governance Framework / Preparedness Framework v2 | “"Under California's Transparency in Frontier AI Act (TFAIA), this FGF is our Frontier AI Framework"; "Under the European Union's General-Purpose AI Code of Practice (the EU CoP), this FGF serves as our publicly available summary of OpenAI's Safety & Security Framework" (§1). PF fn 2: "our adoption of Capability Reports and Safeguards Reports parallels Anthropic's updated RSP."” Also cites {PF} for the fn.2 parallel-to-Anthropic remark. | exact confidence: high | OpenAI Frontier Governance Framework |
| xAI xAI Frontier AI Framework (30 Jun 2026 / 31 Dec 2025) | “June 2026 fn 1: "Terminology used in the The Safety and Security Chapter of the General-Purpose AI Code of Practice developed under the EU AI Act." Dec 2025: claims compliance with TFAIA, Cal. Bus. & Prof. Code 22757.10 et seq.” {FAIF26} PDF metadata reads 'Privileged/Confidential DRAFT working FRAMEWORK DOC'; draft-vs-final status unresolved (open-VERIFY register item 4). Also cites {FAIF25} for the Dec 2025 TFAIA-compliance claim. | exact confidence: medium | xAI Frontier AI Framework (30 Jun 2026) |
| EU EU GPAI Code of Practice (landing page) | “Obligations under unsigned chapters shown through "alternative adequate means" (landing page).” | close confidence: medium | EU GPAI Code of Practice (landing page) |
| California SB 53 California SB 53 | “"A frontier developer that intends to comply with this section by complying with the requirements of, or meeting the standards stated by, a federal law, regulation, or guidance document designated pursuant to subdivision (h) shall declare its intent to do so to the Office of Emergency Services." (22757.13(i)(1)). Also: a system or model card containing the required information is "deemed in compliance" (22757.12(c)(3)).” | exact confidence: high | California SB 53 |
| METR METR | “Common Elements crosswalks lab frameworks to "the EU GPAI Code of Practice and California SB 53".” | close confidence: medium | METR |
| Frontier Model Forum Frontier Model Forum, Risk Taxonomy and Thresholds | “Member synonyms: "Amazon 'Critical Capability Thresholds'; Google 'Critical Capability Levels'; Anthropic 'Capability Thresholds'; OpenAI 'Capability Thresholds'; Microsoft 'Capability Thresholds'" (s3.1).” | close confidence: medium | Frontier Model Forum, Risk Taxonomy and Thresholds |
| Safety Framework Cards (discovery) Safety Framework Cards (SSRN 7061798) | “"9 retrofitted cards" (discovery) [UV].” SFC paywalled/unread; discovery-snippet only. No resolvable primary URL supplied for this tag. | none confidence: low | Discovery sweep (evaluator ecosystem; Safety Framework Cards, unread/paywalled) |
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.
- Anthropic
Explicit non-mapping: "We use this term in its plain meaning rather than adopting any specific statutory definition"; statutory requirements handled "in separate compliance frameworks" (§1 fn 1). Frontier Compliance Framework (SB 53) announced 19 Dec 2025; document not retrieved [UV]. The source table also cites a bracket tag {DAN} for this row that does not resolve to any URL in the supplied sources list -- not carried into the citation.
Anthropic Risk Report (August 2026) - Google DeepMind
FSF fn 1 cites UK DSIT 'Emerging processes', METR, Anthropic RSP and SB 53 compliance framework, OpenAI PF and FMF reports as informing it; the FSF 'does not itself say it is an SB 53 or EU Code of Practice compliance document'.
Google DeepMind Frontier Safety Framework v3.1 - Meta
"we note that there is a lack of consensus among industry and within regulatory frameworks as to how to define some of these terms and concepts." (Appendix I)
Meta Advanced AI Scaling Framework v2
Gap
SB 53's declaration-of-intent record is the only legally operative mapping declaration in the corpus (this is the source's design-implication note, not a literal 'Gap to record' line). NIKOLAI's mapping-declaration registry can copy its shape (declaring party, target instrument, date, revocation) while recording clearly that a NIKOLAI crosswalk entry is analysis, not a declaration by the mapped party.
