OpenAI introduces BioBasicsBench, a 100-task benchmark measuring how well AI agents interpret raw experimental data and handle everyday wet-lab science.

OpenAI built an autonomous pipeline that found, validated, and generated patches for 948 local privilege-escalation vulnerabilities in the Linux kernel.

Discover how ChatGPT and Codex are helping engineers accelerate CAD, PCB design, simulation, and other workflows across the physical world.

Introducing LifeVal, a benchmark measuring how well AI helps people complete everyday tasks, from finding childcare to managing finances and government paperwork.

OpenAI uses layered monitoring to detect frontier model misalignment and unauthorized agent behavior so responders can intervene before incidents escalate.

OpenAI shares model results on the Kakeya conjecture, Exponential Time Hypothesis, and quantum chromodynamics—and how researchers can verify and build on them.

We’re introducing GPT-6.1 Terra, the latest model in the GPT-6 series. GPT-6.1 Terra is a fast, efficient model for high-volume professional work, with strong image analysis and computer use capabilities. This system card addendum provides updates on the baseline safety metrics and evaluation results for GPT-6.1 Terra.

OpenAI explains how we strengthened containment for frontier model training and evaluation, including offline environments and high-compute red-teaming.

See how OpenAI, PerturbAI, and Scripps Research used distributed AI agents to uncover promising research leads in data from millions of brain cells.