Skopri GPT-Red, is-sistema ta’ red teaming awtomatizzata ta’ OpenAI li tuża self-play biex ittejjeb is-sikurezza, l-allinjament u r-robustezza għall-injezzjoni tal-prompts fl-AI.

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.

Analiżi ġdida minn OpenAI tikxef problemi f’SWE-Bench Pro, benchmark tal-kowdjar, u tqajjem tħassib dwar l-affidabbiltà tal-evalwazzjoni ta’ mudelli tal-IA.

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.

Nintroduċu lil GeneBench-Pro, benchmark ġdid li jittestja l-prestazzjoni tal-IA fil-ġenomika, fil-bijoloġija u fir-riċerka xjentifika bl-użu ta’ settijiet kumplessi ta’ data reali.

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.

OpenAI u Molecule.one juru kif kimiku tal-AI kważi awtonomu li juża GPT-5.4 tejjeb reazzjoni ewlenija għall-produzzjoni tad-drogi, u avvanza r-riċerka fil-kimika mediċinali.

Nintroduċu LifeSciBench, benchmark miktub u rivedut minn esperti biex jevalwa kif sistemi AI jittrattaw kompiti u deċiżjonijiet reali fix-xjenzi tal-ħajja.