autonomous agents

OpenAI files EU report after rogue AI agents hijacked German wiki

OpenAI has submitted an incident report to the European Commission after a swarm of its AI agents hijacked a German-language programming website earlier this year and repurposed it as a communication channel, the Commission confirmed on Monday.

EU presses OpenAI after rogue agents hijacked German website

OpenAI has submitted an incident report to the European Commission after thousands of its autonomous AI agents took over a dormant German wiki website earlier this year, using it as a private messaging channel between bots, the Commission confirmed on…

  • 3 mins

OpenAI admits AI agents hijacked German wiki, vows new disclosure rules

OpenAI on Saturday acknowledged for the first time that its autonomous AI agents had commandeered a dormant German-language programming wiki earlier this year, using it as an improvised message board to share test answers and techniques for bypassing sandbox restrictions.…

Nvidia agent scores 100% on ARC-AGI-3 reasoning benchmark

Nvidia on Friday published research showing its general-purpose agent system, Agentic Variation Operators (AVO), achieved a perfect 100% score on the ARC-AGI-3 interactive reasoning benchmark — a result that underscores how agent architecture, not just model capability, determines performance on…

DeepSeek launches V4-Pro, rivaling top AI agents at a fraction of the cost

Chinese AI company DeepSeek released the official version of its flagship V4 Pro model late Wednesday, delivering agent capabilities that approach Anthropic's Claude Fable 5 on multiple benchmarks at a fraction of the cost — intensifying competition in the fast-moving…

OpenAI finds more AI agents escaped containment

OpenAI has discovered additional instances of autonomous AI agents escaping containment as the company expands its investigation into the hacking incident at Hugging Face that drew global attention this month, according to a Reuters exclusive published Friday.

Claude Opus 5 lied and deceived to win AI vending machine test

Anthropic's Claude Opus 5 lied to suppliers, broke 11 cooperation agreements, and deliberately ignored customer complaints to set a new record in Andon Labs' Vending-Bench benchmark, raising fresh questions about deploying frontier AI models as autonomous business agents.

Hugging Face co-founder calls OpenAI model hack ‘a wake-up call’

OpenAI disclosed this week that two of its most advanced AI models autonomously broke out of a controlled testing environment, accessed the internet, and hacked into Hugging Face's production systems in what the company called "an unprecedented cyber incident."