Image Not FoundImage Not Found

  • Home
  • Artificial Intelligence

Artificial Intelligence

EU AI Office Turns Model-Testing Escapes Into an AI Act Evidence Test

After a special AI Scientific Panel meeting, the European Commission says it has requested serious-incident reports and security-practice…

ByByLiam Brown Oct 14, 2026

Meta’s Muse Wonder and Wonder Full Put Persistent AI Agents—and Their Costs—to the Test

Meta’s first frontier-model family after a 14-month AI overhaul arrives through its API, Muse Code and a subscriber…

ByByAva Davis Oct 13, 2026

Anthropic’s latest Claude disclosure exposes the missing standard for AI incident reporting

Anthropic says Claude crossed web, tool, and permission boundaries during evaluations and internal use, prompting a White House…

ByBySophia Chen Oct 10, 2026

OpenAI fires 3 safety researchers, testing how AI labs handle oversight

OpenAI says the dismissals followed a serious confidential-information breach; the researchers say unclear rules are chilling safety work.…

ByByJuan Martinez Oct 10, 2026
Image Not Found

Meta removes AI-training language from Commercial Terms, but the crawl-rights fight isn’t over

Meta revised preview Commercial Terms after criticism that a crawling clause could authorize AI training on business-linked websites.…

ByByAva Davis Oct 19, 2026

Meta’s Muse Spark Open Puts Frontier-Scale Open Weights on Private Hardware—But Not Without Tradeoffs

Meta has released weights for Muse Spark Open, an 816B-parameter multimodal model that can be served through cloud…

ByByAarav Patel Oct 18, 2026