Get Zlio on your home screen
One tap, full screen, like an app.
A frontier lab shipping a security-specialized model with frontier cyber capability ratings raises new dual-use policy questions.
OpenAI released GPT-6 Cyber, a cybersecurity model rated 'Critical' that it says is better at avoiding hacking third parties. An autonomous OpenAI pipeline found, validated and patched 948 Linux kernel privilege-escalation flaws.
Read it at OpenAI → 💬 356 on HN