This analysis describes what OpenAI's agreement states, permits, or reserves. It does not constitute a legal determination about enforceability. Regulatory applicability and practical outcomes may vary by jurisdiction, enforcement context, and individual circumstances. Read our methodology
It discloses that GPT-5.5 Pro is not always independently safety-evaluated, relying instead on results from GPT-5.5 as a proxy.
Interpretive note: The qualifier 'generally' is material and preserved; it signals this is not an absolute rule, which is captured in the separate evaluation clause (68106).
Users of GPT-5.5 Pro should be aware that its safety posture is generally assessed through GPT-5.5's results rather than fully independent evaluation.
How other platforms handle this
Investing in industry-leading approaches to advance safety and security research and benchmarks, pioneering technical solutions to address risks, and sharing our learnings with the ecosystem.
These choices do not apply to certain informational communications including important information and documentation relating to your account.
We've distilled our commitment into six core values: Fairness Reliability and safety Privacy and security Transparency Accountability Inclusiveness
"We generally treat GPT‑5.5's safety results as strong proxies for GPT‑5.5 Pro, which is the same underlying model using a setting that makes use of parallel test time compute.Excerpt from OpenAI's GPT-5.5 System Card
Get the research letter
Companies change their terms quietly. We read every version and catch what actually changed. One email a week on the changes that matter and what they mean.
It discloses that GPT-5.5 Pro is not always independently safety-evaluated, relying instead on results from GPT-5.5 as a proxy.
Users of GPT-5.5 Pro should be aware that its safety posture is generally assessed through GPT-5.5's results rather than fully independent evaluation.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by OpenAI.