Frequency
2
occurrences
First Seen
August 19, 2026
Last Seen
September 2, 2026
Related Threat Clusters
-
OpenAI Launches GPT-5.4-Cyber Amidst Cybersecurity Arms Race
OpenAI has introduced GPT-5.4-Cyber, a specialized AI model for defensive cybersecurity, available only to vetted professionals through its Trusted Access for Cyber (TAC) program. This model is designed to facilitate…
1370 articles · Updated April 14, 2026 -
OpenAI Prepares Astra Model Amid Security Concerns Post-Hugging Face Hack
OpenAI is advancing its Astra model, designed to identify security flaws autonomously, after the Hugging Face hack in July 2026. The Hugging Face incident involved an advanced OpenAI model exploiting a zero-day…
2 articles · Updated September 2, 2026
Recent Intelligence Reports
- OpenAI Teases Astra, Limits Cybersecurity Tools After Hugging Face Hack — Uk.Pcmag · September 2, 2026
- OpenAI slows model training to bolster security after Hugging Face hack — Rnz.Co.Nz · August 19, 2026