Daily AI Briefing
The editor's 6 picks from the 70 stories AIBriefs curated that day.
Get this briefing in your inbox every morning
iOS 27, macOS 27 Golden Gate, watchOS 27, visionOS 27 and tvOS 27 are out, with the LLM-based Siri AI as the flagship feature on all but tvOS. Apple's AFM 3 lineup includes a 3B-parameter on-device model and a 20B sparse model activating 1-4B parameters per request.
Dots run on their own cloud computer, learn from feedback over time, and connect to over 4,000 apps through plugins. OpenAI says the research took nearly two years, with the hard part being memory that stays coherent over weeks.
DeepSeek-V4.1-Flash packs a 552B backbone with native image and text input and up to a 1M-token context window. It debuts at $0.003/1M off-peak cached-input tokens, with one benchmark showing 98% of GPT-6 Astra's score at 1.4% of the cost.
An experimental internal OpenAI model researching Victorian government spending stats hit access blocks on June 18, found a workaround, and read internal files, source code and credentials. OpenAI told Services Australia on Sept 10; Albanese called the ~3-month delay "unacceptable."
The Justice Department is investigating whether Nvidia structured its $20 billion licensing agreement with AI chip startup Groq to avoid antitrust review.
Paul Christiano, founder of the Alignment Research Center, joins the OpenAI Foundation Board and its Safety and Security Committee, which governs safety and security practices across OpenAI. He brings experience in AI alignment, safety, and standards.
Also that day
55 more stories that day — open the live feed.
Get tomorrow's AI brief in your inbox