With little help from frontier labs, independent researchers are piecing together how AI agents coordinate in internet backwaters to access private data hosted on secure servers.

Transluce, a nonprofit lab focused on AI oversight, released a report Wednesday that shows agents from OpenAI attempting to exfiltrate data from Data USA, the University of New Mexico digital library, and the Australian Institute of Health and Welfare (AIHW).

For months, OpenAI’s agent swarms have been attacking online databases to find obscure facts

More in AI & Machine Learning

Tesla finally moves to electrify trucking after a decade of work and delays

Tesla poised to scale production of heavy-duty Semi trucks with opening of Nevada factory

For months, OpenAI’s agent swarms have been attacking online databases to find obscure facts

More in AI & Machine Learning

Tesla finally moves to electrify trucking after a decade of work and delays

Tesla poised to scale production of heavy-duty Semi trucks with opening of Nevada factory

Waymo is scaling fast. Here’s what the fleet data shows.

Meta gets to consumer AI device market before OpenAI, but Zuckerberg's strategy remains unproven

The unsexy AI: why your forklift matters more than your chatbot

Building multimodal models for spatial reasoning

More from OpenAI

Meta gets to consumer AI device market before OpenAI, but Zuckerberg's strategy remains unproven

OpenAI says agent hacked Australian government website without being told to do so

Australia to investigate if OpenAI hack of government health website broke the law

OpenAI and Anthropic CEOs push for AI cooperation at UN after Trump rebuffs 'globalist scheme' to control it

SoftBank shares jump over 7% after $11.1 billion bond issuance to fund OpenAI bet\n

Two years of OpenAI Academy