OpenAI releases new cyber security model to limited group of customers

Chronological Source Flow
Back

AI Fusion Summary

OpenAI has released GPT‑5.4‑Cyber, a more lenient variant of GPT‑5.4, to a limited group of customers, following Anthropic’s Mythos. The model may accept malicious prompts for cybersecurity purposes, unlike its stricter public counterpart.
15/04 01:30 ft.com
3 Πηγές
15/04 11:05 mashable.com
15/04 11:33 dev.to
Comments
Loading...
0