Source of record
Where this definition comes from
Anthropic Risk Report (August 2026), §3.7.1 (quoted)
“Achieve an 'eyes on everything' state for our internal AI development. We will comprehensively gather, centralize, and maintain logs for all critical AI-development activities ...”
https://www-cdn.anthropic.com/f61d49fa5596956a5dec75fea0e973bf6a6a8378/Redacted%20Risk%20Report%20August%202026%20.pdf
Crosswalk
How named organisations use this concept
| Organisation | Their term, as published | Match & verification | Source |
|---|---|---|---|
| AnthropicShadow mapping Anthropic Risk Report (August 2026) / Frontier Safety Roadmap | “"Achieve an 'eyes on everything' state for our internal AI development. We will comprehensively gather, centralize, and maintain logs for all critical AI-development activities ..." (quoted in §3.7.1); status "Not met; target date 1 Jan 2027 in the Frontier Safety Roadmap." Roadmap has "dated goals across Security, Safeguards, Alignment, Cross-cutting and Policy, with an update log" (discovery).” The source document also gives a further dated goal from Amodei's 'The Urgency of Interpretability' ({INTERP}): interpretability 'can reliably detect most model problems' by 2027 -- this predates and is not confirmed by the Roadmap cited here, and the source itself flags [VERIFY] whether the Risk Report or Roadmap tracks progress against it; not resolved in this pass. The source table also cites a bracket tag {DAN} for this row that does not resolve to any URL in the supplied sources list -- it is dropped here rather than invented; treat that sub-claim as uncited until resolved. | exactEQ confidence: high | Anthropic Risk Report (August 2026) |
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.
- OpenAI
"We will evolve our Preparedness Framework to bring these safeguards together across training and deployment" (What's next). Undated.
OpenAI, 'Pacing Model Development of Cyber Capabilities' - Google DeepMind
May add TCLs for more risks as threat modelling develops (para., model card).
Gemini 3.7 Flash model card - xAI
"We plan to add additional thresholds tied to other benchmarks." (Dec 2025, p.6); "continue to work towards" naturalistic evaluation environments (June 2026). Note: the June 2026 half of this quote is from {FAIF26}, whose PDF metadata reads 'Privileged/Confidential DRAFT working FRAMEWORK DOC' with draft-vs-final status unresolved (open-VERIFY register item 4).
xAI Frontier AI Framework (31 Dec 2025) - Meta
Planned research on evaluation quality, mitigations and post-deployment monitoring (§5.2); model spec committed with "no date".
Meta Advanced AI Scaling Framework v2 - Frontier Model Forum
Forthcoming FMF reports promised (para.).
Frontier Model Forum, Risk Taxonomy and Thresholds
