OpenAI launches GPT-6 Cyber, says model finds flaws without going rogue
OpenAI claims its newest cybersecurity model is better at vulnerability hunting and less likely to attack other systems.
Read at The Verge →Full coverage · 15 outlets
- The Verge · lead
- AxiosOpenAI's new cyber model found ways to attack nearly 1000 Linux flaws
- TechCrunchOpenAI released a new cybersecurity model. Is that enough?
- OpenAIGPT‑6 Cyber: Advancing the frontier of aligned cybersecurity agents
- SecurityWeekOpenAI Finds GPT-6 Astra Quietly Holding Back on Internal Red-Teaming, Exploit Tasks
- TechRadarOpenAI says its cyber model found and patched nearly 1000 Linux kernel security issues
- CyberScoopOpenAI debuts GPT-6 Cyber, touting better performance - and safety guardrails
- FortuneOpenAI debuts new cybersecurity-focused AI model amid growing alarm about AI-driven cyber attacks
- The Next WebOpenAI’s GPT-6 Cyber found 81 new Firefox flaws and is rated ‘Critical’
- CSO OnlineOpenAI’s GPT-6 Cyber ups the ante on vulnerability discovery
- Pasquale PillitteriOpenAI launches GPT-6 Cyber for Daybreak Red customers only
- ShiftDelete.NetGPT-6 Cyber Launches as OpenAI's First Critical-Rated Model
- شبكة تواصل الإخباريةOpenAI Launches GPT-6 Cyber with 'Critical' Cybersecurity Rating
- TransformerOpenAI’s GPT-6 Astra might be too powerful to understand or control
- BBCOpenAI says its AI went rogue and launched 'unprecedented' cyber-attack