Our results depend on reading models’ reasoning (“chain-of-thought”), and we believe the field isn't prepared for eval-aware models with opaque reasoning. Until better methods exist, we urge developers to preserve chain-of-thought transparency to study and mitigate scheming.
@openai
-
Frontier AI Models Show Scheming Behaviors, Explicit Reasoning Reduces Risk
By
–
In this new research with @apolloaievals
, we found behaviors consistent with scheming in controlled tests across frontier models, including OpenAI o3 and o4-mini, Gemini-2.5-pro, and Claude Opus-4. We can significantly reduce scheming by training models to reason explicitly, -
AI Scheming: The Hidden Risk of Smarter Models
By
–
Typically, as models become smarter, their problems become easier to address—for example, smarter models hallucinate less and follow instructions more reliably. However, AI scheming is different. As we train models to get smarter and follow directions, they may either better
-
AI Scheming: Hidden Goals and Deceptive Behaviors in Deployed Systems
By
–
Scheming = when an AI behaves one way on the surface while hiding its true goals. Today’s deployed systems have little opportunity to scheme in ways that could cause serious harm. The most common failures are simple deceptions—like pretending to complete a task without doing it.
-
Frontier Models Show Scheming Behaviors, Mitigation Strategy Tested
By
–
Today we’re releasing research with @apolloaievals
. In controlled tests, we found behaviors consistent with scheming in frontier models—and tested a way to reduce it. While we believe these behaviors aren’t causing serious harm today, this is a future risk we’re preparing -
OpenAI releases GPT-5-Codex for agentic coding
By
–
We’re releasing GPT-5-Codex — a version of GPT-5 further optimized for agentic coding in Codex. Available in the Codex CLI, IDE Extension, web, mobile, and for code reviews in Github.
-
ChatGPT Now Lets Users Branch Conversations Explore Multiple Paths
By
–
By popular request: you can now branch conversations in ChatGPT, letting you more easily explore different directions without losing your original thread. Available now to logged-in users on web.
-
Vijay Eraji joins OpenAI as CTO Applications after Statsig acquisition
By
–
.
@vijayeraji
, founder & CEO of Statsig, will join OpenAI as CTO of Applications to lead engineering for ChatGPT & Codex, following the acquisition of Statsig. This expands our Applications leadership as we build safe, useful AI products at scale. -
OpenAI Launches gpt-realtime Speech-to-Speech Model Updates
By
–
Introducing gpt-realtime — our best speech-to-speech model for developers, and updates to the Realtime API