Examples
Examples from published documents
- The EU Code of Practice lists "staging the access to the model, e.g. by limiting API access to vetted users, gradually expanding access based on post-market monitoring, and/or not making the model parameters publicly available for download initially" as a safety mitigation (EU GPAI Code of Practice, Safety and Security Chapter, Measure 5.1)
- Anthropic reports that "some launch preview customers had early access using the less robust classifier until we developed an improved classifier for the model's general access launch" (Anthropic Risk Report, August 2026, §6.5)
Counter-examples
Not to be confused with
- An exemption gives a named group reduced safeguards on a standing, revocable basis; a release stage gives a limited group access before others, normally with the same or stronger safeguards.
Source of record
Where this definition comes from
EU GPAI Code of Practice, Safety and Security Chapter, Measure 5.1, example (4)
“staging the access to the model, e.g. by limiting API access to vetted users, gradually expanding access based on post-market monitoring, and/or not making the model parameters publicly available for download initially”
https://ec.europa.eu/newsroom/dae/redirection/document/118119G42 Frontier Safety Framework, §3, Deployment Mitigation Levels 2 and 3
“Controlled Rollout: For new frontier-level models, implement phased rollouts, starting with limited access to trusted users, with full deployment only after exhaustive risk assessments.”
https://www.g42.ai/application/files/9517/3882/2182/G42_Frontier_Safety_Framework_Publication_Version.pdf
Crosswalk
How named organisations use this concept
Every row below is a shadow mapping. A shadow row is CASRAI's own reading of a published document. No lab, evaluator or regulator named on a shadow row has declared, endorsed, or been consulted on it. That changes only when an organisation files its own Mapping Declaration.
| Organisation | Their term, as published | Match & verification | Source |
|---|---|---|---|
| EUShadow mapping EU GPAI Code of Practice, Safety and Security Chapter | “Measure 5.1 safety mitigation: "staging the access to the model, e.g. by limiting API access to vetted users, gradually expanding access based on post-market monitoring, and/or not making the model parameters publicly available for download initially"” | exactEQ confidence: high | EU GPAI Code of Practice, Safety and Security Chapter |
| G42Shadow mapping G42 Frontier Safety Framework | “"Controlled Rollout: For new frontier-level models, implement phased rollouts, starting with limited access to trusted users, with full deployment only after exhaustive risk assessments."” | exactEQ confidence: high | G42 Frontier Safety Framework |
| OpenAIShadow mapping OpenAI Preparedness Framework v2 | “Trust-based Access safeguards: "Iterated deployment to wider user groups"; efficacy assessment: "Iterated deployment to wider user group, with careful monitoring to detect instances of abuse"” | closeCL confidence: high | OpenAI Preparedness Framework v2 |
| MetaShadow mapping Meta Advanced AI Scaling Framework v2 | “Deployment types: "Limited deployment: releasing a final version externally to a limited set of vetted and trustworthy external parties."; "Controlled deployment: deploying to a small percentage-limited set of product users prior to broad availability."” Meta's list is a deployment-type ladder; the staging is implied by the order, not stated as a rule. | closeCL confidence: medium | Meta Advanced AI Scaling Framework v2 |
| Google DeepMindShadow mapping Frontier Safety Framework v3.1 | “"Low-Risk External Deployments: represent small scale and private releases restricted to particular external groups or organizations."; residual risk assessment considers that "small scale and private deployments may pose substantially less risk than large scale or public deployments"” | closeCL confidence: medium | Google DeepMind Frontier Safety Framework v3.1 |
| AnthropicShadow mapping Anthropic Risk Report, August 2026 | “"we additionally piloted a staged deployment process in which we first deployed the model on internal surfaces with stronger blocking controls against dangerous actions to gather more real-world usage data in a lower-risk setting before rolling it out for unrestricted internal deployment" (§2.18)” Staging applied to internal deployment only. | narrowNR confidence: medium | 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
