Bug Bounty Researchers Hacked OpenAI Using Anthropic's Claude - DTH
Summary
Host Sarah Lane covers a security breach where Hackron AI used Claude to exploit OpenAI’s forum, alongside unsealed court filings revealing internal warnings about AI scraping. The episode details Lucid and Bolt’s plan for 25,000 European robotaxis, a German court ruling holding Meta liable for fake ads, and the SEC’s new exemption for tokenized stocks. Additionally, it reports on OpenAI’s launch of a legal research tool, expanded Costco delivery partnerships, and ongoing legal disputes involving Apple and Elon Musk.
Topics discussed
Sponsor: Capital One's Multi-Agent AI
Sponsor: ShipStation for Peak Season
Hackron AI exploits OpenAI forum bug
Internal warnings on AI scraping in court filings
Lucid and Bolt plan European robot taxi fleet
German court rules Meta liable for fake ads
SEC exemption for tokenized stock trading
OpenAI launches Astra for Law
Costco expands delivery via DoorDash and Uber
OpenAI expected to solve Hodge Conjecture
Judge denies OpenAI access to Apple-Musk deal
Show notes and sign-off
Sponsor: Capital One's Multi-Agent AI
Sponsor: ShipStation Order Fulfillment
Listen ad-free on Castria