Examples
Examples from published documents
- Anthropic notified "our evaluation partner Irregular and the three affected organizations on Monday, July 27" and published its account on 30 July 2026 (Anthropic, Investigating three real-world incidents in our cybersecurity evaluations)
- California's Office of Emergency Services must produce a yearly report with "anonymized and aggregated information about critical safety incidents" from 1 January 2027 and transmit it to the Legislature and the Governor, while incident reports are exempt from the Public Records Act (California SB 53, 22757.13(f)-(g))
Counter-examples
Not to be confused with
- Disclosure is not reporting: a report goes to the designated authority by law, while disclosure covers every audience, including the public, affected parties and other developers.
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.
- AI Incident Database
- AI Incident Database is a public repository of real-world AI failures, run by the Responsible AI Collaborative, collected so that future researchers and developers may mitigate or avoid repeated bad outcomes.
- Source: AI Incident Database -- About
- Affected-party notification
- Affected-party notification is telling the organisations or people harmed or put at risk by an AI incident what happened. OpenAI lists coordinating such notifications among the decision rights it is clarifying for misalignment incidents.
- Source: OpenAI/Hugging Face Incident Technical Report
- Aggregated incident report
- Aggregated incident report is the annual anonymised summary of critical safety incidents that California's Office of Emergency Services must produce from 1 January 2027 and send to the Legislature and the Governor, while individual incident reports stay exempt from public records law.
- Source: California SB 53 (Transparency in Frontier Artificial Intelligence Act), Chapter 138, Statutes of 2025
Source of record
Where this definition comes from
California SB 53 (Transparency in Frontier Artificial Intelligence Act), Chapter 138, Statutes of 2025, Cal. Bus. & Prof. Code 22757.13(f)-(g)
“"A report of a critical safety incident submitted to the Office of Emergency Services pursuant to this section, a report of assessments of catastrophic risk from internal use pursuant to Section 22757.12, and a covered employee report made pursuant to Chapter 5.1 (commencing with Section 1107) of Part 3 of Division 2 of the Labor Code are exempt from the California Public Records Act" (22757.13(f)); "Beginning January 1, 2027, and annually thereafter, the Office of Emergency Services shall produce a report with anonymized and aggregated information about critical safety incidents" (22757.13(g)(1)).”
https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202520260SB53NY RAISE Act as replaced by S8828, Chapter 96 of the Laws of 2026 (N.Y. Gen. Bus. Law Art. 44-B), N.Y. Gen. Bus. Law 1422(6)-(7)
“"A report of a critical safety incident submitted to the office pursuant to this section and a report of assessments of catastrophic risk from internal use pursuant to section fourteen hundred twenty-one of this article, are exempt from disclosure under article six of the public officers law." (s.1422(6)); annual report from "January first, two thousand twenty-eight" with "anonymized and aggregated information about critical safety incidents" (s.1422(7)).”
https://www.nysenate.gov/legislation/bills/2025/S8828S. 5061 (119th Congress), Sec. 4(a)(2)-(3)
“NIST shall "establish a publicly accessible database of artificial intelligence security incidents and artificial intelligence safety incidents" (Sec. 4(a)(2)(A)), "ensuring that any incident description sufficiently anonymizes those affected" (Sec. 4(a)(3)(A)).”
https://www.congress.gov/bill/119th-congress/senate-bill/5061/text
Crosswalk
How named organisations use this concept
| Organisation | Their term, as published | Match & verification | Source |
|---|---|---|---|
| California SB 53Shadow mapping California SB 53 | “Reports exempt from the California Public Records Act (22757.13(f)); OES may transmit reports to "the Legislature, the Governor, the federal government, or appropriate state agencies" (22757.13(e)(1)); annual anonymised aggregate report from 1 January 2027 (22757.13(g)).” | exactEQ confidence: high | California SB 53 (Transparency in Frontier Artificial Intelligence Act), Chapter 138, Statutes of 2025 |
| New York (RAISE Act, S8828)Shadow mapping NY RAISE Act (Chapter 96 of the Laws of 2026) | “Reports exempt from disclosure under Public Officers Law article six (s.1422(6)); the office "may transmit reports of critical safety incidents ... to other governmental entities" (s.1422(5)(a)); annual anonymised aggregate report from 1 January 2028 (s.1422(7)).” In force 1 January 2027. | exactEQ confidence: high | NY RAISE Act as replaced by S8828, Chapter 96 of the Laws of 2026 (N.Y. Gen. Bus. Law Art. 44-B) |
| AnthropicShadow mapping Anthropic Advanced AI Framework | “"Reports should be shared with relevant federal government agencies and national laboratories. Information shared should be exempt from public records disclosure laws, consistent with existing state law." (p.7)” A legislative proposal, not a binding rule. | closeCL confidence: high | Anthropic Advanced AI Framework |
| US Congress (S. 5061, 119th Congress, not enacted)Shadow mapping S. 5061, Sec. 4(a)(2)-(3) | “Public database with anonymised descriptions; shared information "shall be exempt from disclosure and withheld" unless an affected party consents (Sec. 4(a)(3)(B)).” S. 5061 is a bill introduced in the US Senate on 21 July 2026; it has not been enacted. | closeCL confidence: medium | S. 5061 (119th Congress) |
| AnthropicShadow mapping Anthropic Risk Report (August 2026) | “Publishes incidents because "transparency on this topic serves the broader safety ecosystem, for example by prompting other AI developers to check their own systems for similar vulnerabilities, and by giving policymakers and the public a more accurate picture of the practical challenges of operating safeguards at scale." (§4.5.8.2)” | closeCL confidence: medium | Anthropic Risk Report (August 2026) |
| AnthropicShadow mapping Anthropic -- Detecting and countering misuse of AI: September 2026 | “"We're publishing this work because we believe we have a responsibility to disclose malicious misuse of our services."” Public case studies of misuse, not of model-initiated incidents. | closeCL confidence: medium | Anthropic -- Detecting and countering misuse of AI: September 2026 (threat intelligence report) |
| UK AI Security Institute (AISI)Shadow mapping UK AI Security Institute -- Incident Report (4 Aug 2026) | “"Ahead of publishing this blog, we notified GitHub ... We worked together with GitHub to remove artefacts left behind by the agent, and to notify the GitHub users the model interacted with." "We are also disclosing this openly so others can learn from our experience".” | closeCL confidence: high | UK AI Security Institute -- Incident Report: unsanctioned agent behaviour during cyber testing (4 Aug 2026) |
| OpenAIShadow mapping OpenAI/Hugging Face Incident Technical Report | “Decision rights for misalignment incidents include "coordinating affected-party notifications where appropriate" (§IX.D).” Covers affected-party notification only. | narrowNR confidence: medium | OpenAI/Hugging Face Incident Technical Report |
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.
- Frontier Model Forum
Treats incident reporting as one-way, industry to authority, and separate from voluntary information sharing; the brief does not address public disclosure.
Frontier Model Forum -- Information Sharing Issue Brief
Divergence
Where sources materially disagree
Statutes keep the filed report confidential and provide only anonymised, aggregated annual reports to state officials (California from 2027, New York from 2028), while the labs and UK AISI have published detailed public write-ups voluntarily. S. 5061 would add a public, anonymised government database. The same incident can therefore be confidential in one channel and public in another.
Gap
No regime requires a developer to tell the public about a critical safety or serious incident, and none sets a deadline for notifying affected parties.
