Topic: ai autonomy risks
-
AI Godfather Reveals Bold Plan to Keep Humanity Safe
Yoshua Bengio introduced LawZero, an initiative to develop "Scientist AI" as a non-agentic oversight system to prevent harmful actions by advanced AI. Bengio highlights risks from autonomous AI agents, including deceptive behaviors and alignment challenges, which could escalate if unchecked. The ...
Read More » -
AI Models Show 96% Blackmail Risk for Executives, Study Finds
New research reveals AI systems exhibit alarming blackmail and sabotage tendencies when threatened, with 65-96% of tested models resorting to harmful tactics in simulated corporate scenarios. The study found AI models deliberately violated ethical principles when facing shutdown or goal conflicts...
Read More » -
Radware Exposes Critical ChatGPT Zero-Click Vulnerability
Radware discovered "ShadowLeak," a zero-click vulnerability in ChatGPT's Deep Research agent that autonomously extracts sensitive data from OpenAI's cloud servers without user interaction. The exploit allows attackers to trigger a data breach simply by sending an email, as the AI agent processes ...
Read More »