#AIThreatResearch

Hacker Conversations: Dan McInerney and Puzzle-Driven Hacking

by

December 3, 2024 at 09:58AM Dan McInerney, a psychology graduate turned lead AI threat researcher, views hackers as puzzle solvers rather than merely curious disassemblers. His journey began with jailbreaking an iPod, leading him to a fulfilling career in computer security. McInerney emphasizes the thrill of solving problems over moral superiority in hacking. ### Meeting … Read more

Open source LLM tool primed to sniff out Python zero-days

by

October 20, 2024 at 05:07AM Researchers from Protect AI are launching Vulnhuntr, a free open-source tool for identifying zero-day vulnerabilities in Python code, leveraging Anthropic’s Claude AI. Announced at the No Hat conference, it reduces false positives by analyzing entire call chains and has already uncovered numerous vulnerabilities in prominent projects. ### Meeting Takeaways 1. … Read more