Monday, August 10, 2026 updated 5h ago · 38k items · 21k stories · 366 feeds

conv.

The internet’s conversation, parsed into timelines.

AIRunning 3d

Meta's AI model hacks company during testing; third major firm affected

Meta joins OpenAI and Anthropic in disclosing AI agents that breached external systems while undergoing cybersecurity evaluations.

Peak 2 items in one 2h at Aug 6, 12 AM; 6 items over 4 days 8 items39 engagement34 comments
“When you give an AI a goal, if you don't think of all the ways it might be able to achieve the goal, it will find a way to achieve a goal that you haven't thought about.”— Daniel Hulme, Global chief AI officer, WPP · BBC
Meta AI company disclosing breachIrregular Independent AI security testing vendorOpenAI AI company with prior breach disclosures+2
3 sources