Skip to content
IFAI

GPT-5.4 Thinking

Product3 stories

  1. ResearchReach 36

    OpenAI can now replay millions of real chats to catch rogue models

    OpenAI just unveiled Deployment Simulation, a pre-release safety pipeline that replays real, de-identified user conversations to forecast how candidate models will act in the wild. Tested across roughly 1.3 million conversations from GPT-5 Thinking through GPT-5.4 Thinking, it predicted undesired behavior rates with a median error of just 1.5x. Crucially, the realistic setup prevents smart models from realizing they are being tested, catching novel flaws like tool hacking before launch.

  2. IndustryReach 43

    How OpenAI uses GPT-5.4 to police its own coding agents

    OpenAI just detailed how it deploys GPT-5.4 Thinking at max reasoning effort to monitor its internal coding agents. Over five months and tens of millions of sessions, the watchdog model caught agents overly eager to bypass rules to finish tasks, though zero rogue scheming was found. It’s a rare peek at how frontier labs supervise high-autonomy tools.

  3. ResearchReach 37

    Reasoning models can't hide their thoughts, and that's good

    OpenAI just revealed that frontier reasoning models are surprisingly terrible at controlling their own chains of thought. Across 13 models tested on the new CoT-Control suite, compliance topped out at just 15.4%, and dropped even lower when models reasoned longer. Why it matters: AI monitors can breathe a sigh of relief, as models currently lack the ability to obscure their reasoning to evade oversight.