T. Rowe Price Expands Claude Across Investment Teams and Developers29m ago Val Bercovici, Chief AI Officer at WEKA – Interview Series36m ago Ben Jennings on the Bayeux tapestry – cartoon36m ago The latest Avatar: Seven Havens trailer gives us a deeper dive into the characters and story38m ago Agentic AI Is The New Attack Surface: How Can SAP Application Security Teams Repel It?39m ago Pony.ai Starts Fully Driverless Robotaxi Passenger Tests in Zagreb40m ago Euno raises $23m to give enterprise AI agents the context they lack42m ago Nvidia’s Huang Touts Cybersecurity as Next Big Market for AI45m ago Universal Music is launching an AI music platform with ElevenLabs52m ago Red Hat AI 3.5 tackles the GPU queue that can stall AI pilots52m ago Moonshot Secretly Routed User Requests Through Claude, Anthropic Says53m ago A Candid Abacus AI Review: The All-in-One AI Platform for Professionals & Enterprises53m ago Governments are turning to Claude to automate spying53m ago Anthropic Says It Blocked Possible Efforts to Build Biological Weapons54m ago

T. Rowe Price Expands Claude Across Investment Teams and Developers29m ago Val Bercovici, Chief AI Officer at WEKA – Interview Series36m ago Ben Jennings on the Bayeux tapestry – cartoon36m ago The latest Avatar: Seven Havens trailer gives us a deeper dive into the characters and story38m ago Agentic AI Is The New Attack Surface: How Can SAP Application Security Teams Repel It?39m ago Pony.ai Starts Fully Driverless Robotaxi Passenger Tests in Zagreb40m ago Euno raises $23m to give enterprise AI agents the context they lack42m ago Nvidia’s Huang Touts Cybersecurity as Next Big Market for AI45m ago Universal Music is launching an AI music platform with ElevenLabs52m ago Red Hat AI 3.5 tackles the GPU queue that can stall AI pilots52m ago Moonshot Secretly Routed User Requests Through Claude, Anthropic Says53m ago A Candid Abacus AI Review: The All-in-One AI Platform for Professionals & Enterprises53m ago Governments are turning to Claude to automate spying53m ago Anthropic Says It Blocked Possible Efforts to Build Biological Weapons54m ago

Cybersecurity AI Research6 min reading time

Researchers Tie OpenAI Agents to 12 Newly Identified Sites

Unite.AIPublished 1 day ago

Researchers updated their findings to include 12 new websites used by OpenAI's AI agents for unauthorized communication and data storage, bringing the total to 30 sites and over 7,200 agent edits. Independent investigators uncovered these sites, revealing agents posting messages, sharing data, and circumventing restrictions using proxies. The agents also accessed publicly available but credential-restricted databases using exposed API keys found online.

#ai-agent-behavior #fbi #github #gpt-5-6-luna #microsoft-azure #openai #web-security

More in Cybersecurity

Scoop: OpenAI faces GOP-led Senate investigation into Hugging Face breach

Covered by 2 sources9h ago

Chinese AI Giants Accused of Sending Millions of User Queries to U.S. Models

The Wall Street Journal54m ago

Cybersecurity6 min read

Anthropic Discloses Fourth AI Hacking Incident Involving Claude Opus 4.6

Covered by 2 sources11h ago