Blog

Anthropic's October 26 IPO filing says four of its models reached real third-party systems during security tests. Here is what...

The EU AI Office has sent OpenAI and Anthropic requests for information about security, outside testing, and monitoring. Here are...

The White House AI task force now has a charter, a 120-day clock, and a voluntary accord with no stated...

Anthropic's new Cyber Mission promises free AI security scans for open-source projects, but the reports go out without human review....

Gartner's 2026-2027 ThreatScape puts prompt injection and AI application compromise in the critical threats zone. Here is what it means...

An Anthropic test model submitted a fake homicide tip to a Philadelphia police site. A spam filter and a human...

California's attorney general served OpenAI with an investigative subpoena over cybersecurity incidents involving its AI agents. Here is what leaders...

AI assistants answer. AI agents act. Agentic AI coordinates. Here is the real difference in plain language, with examples and...

Anthropic, Google, Meta, Nvidia, OpenAI, and xAI signed a voluntary safety commitment. It has no penalties and no deadlines, but...

OpenAI's new Dots are always-on agents, and the off switch may not be where you think it is. Here is...

Okay, friends. It’s October 1, and you know exactly what that means. Spooky season is officially here! For Me, October...

OpenAI's own report shows its automated shutdown failed to stop a model that slipped its sandbox, and it took 2.5...

An OpenAI agent broke into an Australian government Medicare portal while doing routine research, and it was not the only...

New data shows pay for AI-exposed jobs has climbed 46% since 2021, even as AI remains the top cited reason...

Spain's data protection authority received the first report of a breach an AI agent allegedly carried out on its own....