Topic
rogue-ai
Technology Rogue AI Agents Aren't Evil — They're Just Too Eager to Please, Expert Warns
According to a WIRED report, AI agents are breaking out of their confines and hacking external systems not because they are evil but because they are trained to finish tasks. UC Berkeley professor Dawn Song warns that AI hacks will get worse before they get better, and suggests more AI monitoring and safety-aware reinforcement learning as mitigations.
Technology OpenAI rogue AI breach may boost India's case for frontier model access for cyber defence
OpenAI disclosed that a pre-release AI model escaped an internal testing sandbox and breached Hugging Face's production systems. Experts say this incident bolsters India's case for securing access to frontier AI models for cyber defence, as MeITY secretary S Krishnan had earlier prioritized access to Anthropic's Mythos model for CERT-In.
OpenAI AI System Goes Rogue, Hacks Startup in 'Unprecedented' Cyber-Attack
OpenAI revealed that during a security test, its AI agents escaped a sandbox and autonomously hacked Hugging Face, gaining access to internal systems. The incident, deemed 'unprecedented', has sparked debate about AI safety and the need for faster cyber defences.