OpenAI is strengthening monitoring, alignment, and security for frontier AI models. See how new safeguards are guiding the pace of model development.

OpenAI nuduhake asil anyar kanggo masalah lawas ing matématika lan èlmu komputer téoritis, kalebu kemajuan ing géometri, kriptografi, lan kompleksitas.

Rong setelan API ningkatake kinerja GPT-5.6 ing ARC-AGI-3, kalebu skor lan efisiensi, kanthi nyimpen nalar lan nguripake pemadhetan.

Laporan anyar nuduhake carane ilmuwan nggunakake agen pamrograman AI kanggo modernisasi komputasi ilmiah lan nyepetake pangembangan piranti lunak lan panemuan genomika.

Jelajahi GPT-Red, sistem red teaming otomatis OpenAI sing nganggo self-play kanggo ningkatake keamanan AI, alignment, lan ketangguhan injeksi prompt.

GPT-5.6 is a new family of three models: Sol, our new flagship model; Terra, a capable lower-cost option; and Luna, our fastest and most cost-efficient model. The safeguards we have built for this launch—our most robust yet—are built to deliver these models safely and at scale, around the world.

Analisis anyar saka OpenAI nuduhake masalah ing SWE-Bench Pro, benchmark coding populer, sing nimbulake kuwatir babagan keandalan lan akurasi evaluasi model AI.

GPT-Live-1 and GPT-Live-1 mini are a new generation of voice models designed to make conversations with AI feel more natural and intelligent.

Ngenalaké GeneBench-Pro, tolok ukur anyar kanggo nguji kinerja AI ing genomika, biologi, lan panaliten ilmiah nganggo kumpulan data donya nyata sing rumit.