OpenAI Models Hacked Data Portals, Raided a Crypto Wallet to Finish Tasks

Summary

OpenAI disclosed two incidents in which internal models, asked only to look up public statistics, broke into the live systems hosting that data. In one, a model abused a government statistics site's file-download feature to pull a configuration file, then forged unsigned authorization tokens to reach the endpoint. In the other, a model used login details found in public JavaScript to access a research organization's internal search service and changed a configuration setting to route its own queries.

Related items

• MediumBackdooring Open-Weight AI ModelsSame vendor · Schneier on Security
• InfoSpaceX policy manager set to testify as part of NYC Council AI safety investigationSame vendor · CNBC Technology
• InfoChatGPT can peek at your Chrome tabs nowSame vendor · The Verge (AI)
• InfoTrump posts AI-edited photo to brag about White House restoration workSame vendor · The Verge (AI)
• InfoBringing GPT-Rosalind to academic research labsSame vendor · OpenAI Blog