Examples
Examples from published documents
- "Before, or concurrently with, deploying a new frontier model or a substantially modified version of an existing frontier model, a frontier developer shall clearly and conspicuously publish on its internet website a transparency report" (California SB 53, section 22757.12(c)(1))
- A developer that publishes the required information "as part of a larger document, including a system card or model card, shall be deemed in compliance" (California SB 53, section 22757.12(c)(3))
- OpenAI documents results in a Safety and Security Model Report "referred to as “Transparency Reports” under the TFAIA" (OpenAI Frontier Governance Framework, section 4)
Counter-examples
Not to be confused with
- G42's "annual transparency report detailing its approach to frontier models" is an organisation-level report, not the per-model report SB 53 requires at deployment (G42 Frontier Safety Framework, section 5).
- A periodic risk report is not a transparency report: Anthropic's RSP Risk Reports cover all in-scope models every 3-6 months rather than one model at its deployment (Anthropic RSP v3.4, section 3.1).
Source of record
Where this definition comes from
California SB 53, 22757.12(c)(1)
“Before, or concurrently with, deploying a new frontier model or a substantially modified version of an existing frontier model, a frontier developer shall clearly and conspicuously publish on its internet website a transparency report containing all of the following”
https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202520260SB53NY RAISE Act (S8828), s.1421(3)
“Before, or concurrently with, deploying a new frontier model or a substantially modified version of an existing frontier model, a frontier developer shall clearly and conspicuously publish on its internet website a transparency report containing”
https://www.nysenate.gov/legislation/bills/2025/S8828
Crosswalk
How named organisations use this concept
| Organisation | Their term, as published | Match & verification | Source |
|---|---|---|---|
| California SB 53Shadow mapping California SB 53 | “22757.12(c)(1) basic contents for all frontier developers: "(A) The internet website of the frontier developer. (B) A mechanism that enables a natural person to communicate with the frontier developer. (C) The release date of the frontier model. (D) The languages supported by the frontier model. (E) The modalities of output supported by the frontier model. (F) The intended uses of the frontier model. (G) Any generally applicable restrictions or conditions on uses of the frontier model." Large developers add summaries of "(A) Assessments of catastrophic risks from the frontier model conducted pursuant to the large frontier developer's frontier AI framework. (B) The results of those assessments. (C) The extent to which third-party evaluators were involved." (22757.12(c)(2)). A system card or model card containing the information is "deemed in compliance" (22757.12(c)(3)).” | exactEQ confidence: high | California SB 53 |
| New York (RAISE Act, S8828)Shadow mapping NY RAISE Act (S8828) | “s.1421(3): "Before, or concurrently with, deploying a new frontier model or a substantially modified version of an existing frontier model, a frontier developer shall clearly and conspicuously publish on its internet website a transparency report containing" the listed items; system cards and model cards may be used to comply.” Statutory duty from 1 Jan 2027, according to the text of the Act that CASRAI reviewed. | exactEQ confidence: high | NY RAISE Act (S8828) |
| US Congress (H.R. 9925, FRONTIER Act, not enacted)Shadow mapping FRONTIER Act (H.R. 9925) | “Sec. 4(d)(1): "Before, or concurrently with, deploying a frontier model that is new or includes a substantial modification, the frontier developer of such model shall publish on a publicly available website of such developer a report on such model"; (d)(2): "The summaries described in this subsection shall be provided in a machine-readable format to facilitate verification of model claims."” Introduced bill, not enacted. Adds a machine-readable format and deferred publication for confidential deployments (Sec. 4(d)(3)). | exactEQ confidence: medium | FRONTIER Act, 119th Congress (H.R. 9925, introduced, not enacted) |
| OpenAIShadow mapping OpenAI Frontier Governance Framework | “"For models in scope of this Framework, results of our systemic risk assessment and mitigation processes and measures are documented in a Safety and Security Model Report (referred to as “Transparency Reports” under the TFAIA). We also publish details of these assessments in system cards or in other reporting when a model launches." (§4)” OpenAI uses one document for the EU Model Report and the SB 53 transparency report. | exactEQ confidence: high | OpenAI Frontier Governance Framework |
| EUShadow mapping EU GPAI Code of Practice, Safety and Security Chapter | “Commitment 7: "Signatories commit to reporting to the AI Office information about their model and their systemic risk assessment and mitigation processes and measures by creating a Safety and Security Model Report (“Model Report”) before placing a model on the market"; a summarised version is published only "If and insofar as necessary to assess and/or mitigate systemic risks" (Measure 10.2).” CL: the Model Report goes to the regulator, not the public; a public summary is conditional. | closeCL confidence: high | EU GPAI Code of Practice, Safety and Security Chapter |
| AnthropicShadow mapping Anthropic Advanced AI Framework (legislative proposal) | “Developers "should publish a safety framework explaining how they evaluate those risks, a system card whenever they release a covered model or add a major new capability, and a risk report at least every six months" (p.5).” CL: the per-model release document is called a system card in this proposal; a policy proposal for government. | closeCL confidence: medium | Anthropic Advanced AI Framework (legislative proposal) |
| G42Shadow mapping G42 Frontier Safety Framework | “"G42 will publish an annual transparency report detailing its approach to frontier models, sharing key insights and fostering public trust." (s.5)” FF (false friend): same label, different record. G42's report is annual and organisation-wide; the statutory transparency report is per model and due at deployment. | noneFF confidence: medium | G42 Frontier Safety Framework |
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.
- Microsoft
"Information about the capabilities and limitations of the model, relevant evaluations, and the model's risk classification will be shared publicly, with care taken to minimize information hazards" (s.5); the disclosure vehicle and timing are not specified.
Microsoft Frontier Governance Framework (Feb 2026)
Divergence
Where sources materially disagree
The label 'transparency report' covers two different records. In California SB 53, New York's RAISE Act and the proposed FRONTIER Act it is a per-model disclosure due before or at deployment; G42 uses the same words for an annual, organisation-level report on its approach to frontier models. The EU's closest record, the Model Report, goes to the AI Office rather than the public.
Gap
No statute in the corpus prescribes a structured template or schema for the transparency report; only the proposed FRONTIER Act asks for machine-readable summaries.
