Katie Paxton-Fear, Tabletop Scenarios, cje, and Brendan Dolan-Gavitt highlight AI risks from unsafe command approvals and data leakage to unresolved defender-dilemma debates and models confidently reacting to their own mistakes.
This permission model of “ask the human before running commands” worries me
That AI agent you deployed that trained on all of your internal corporate documents is now spilling tea to anyone who chats with it.
AI's greatest contribution to cybersecurity may end up being that it finally settles the debate over whether the defender's dilemma is actually real.
like 60% of the time its excitement is about a mistake it just made
This finding is one of many signals tracked across Cyber Security. The live feed updates every few hours with new authority voices, debates, and emerging ideas.
← Back to Cyber Security