Google's Gemma 4 runs entirely offline on iPhone while Adobe's Firefly AI can now control entire Creative Cloud workflows.

Anthropic launches self-managing code routines, investors question OpenAI's massive valuation, and Google integrates AI prompts into everyday browsing.

We dive into OpenAI's leaked strategy memo revealing their new 'Spud' model, plus the industry-wide compute crisis that's driving GPU prices up 50%.

Arcee AI challenges Claude with a massive open model, Berkeley researchers reveal AI benchmarks are broken, and ChatGPT faces a stalking lawsuit.

Google's Gemma 4 brings agentic AI to your phone without cloud connections, Berkeley exposes flaws in AI benchmarks, and OpenAI continues its acquisition spree.

A Molotov cocktail targets OpenAI's CEO, Linux officially embraces AI coding, and the CIA produces its first fully autonomous AI intelligence report.

OpenAI cuts Pro pricing by 50% to compete with rivals, while facing investigation after ChatGPT was reportedly used to plan a deadly campus attack.

Meta's shocking shift to proprietary AI with Muse Spark, plus a breakthrough that lets you train 100B+ models on just one GPU.

Breaking down Anthropic's cybersecurity-focused Claude Mythos model and their multi-gigawatt compute deal, plus AI rivals unite against model copying.

OpenAI proposes radical economic changes for the AI age, while new research reveals AI cyber capabilities are doubling every six months.