THE DAILY BRIEF

According to the Financial Times, OpenAI disbanded its preparedness team at the end of last month, eliminating the group responsible for assessing whether models posed serious risks and developing mitigation strategies.

► The move signals OpenAI deprioritizing internal safety oversight just as its models approach more powerful capabilities, raising questions about risk management at the industry leader.

OpenAI released a ChatGPT variant designed specifically for young users and plans to automatically enroll all young users in the new experience.

► The auto-enrollment strategy extends AI dependency to younger audiences at scale, amplifying the stakes of the Meta addiction trial and raising questions about consent and age-appropriate AI design.

Anthropic's market share for business AI model usage reached 43.5% in July, extending its lead over OpenAI, while newer models like Fable 5 struggle to gain adoption due to high costs.

► Anthropic's dominance in enterprise token consumption—not just startup mindshare—signals a structural shift in which AI provider will control the most valuable customer relationships.

A video discovered in the macOS Tahoe 26.7 release candidate shows camera-equipped AirPods using Visual Intelligence, indicating the hardware is nearly ready to launch.

► Camera AirPods represent Apple's next frontier in wearable AI and visual search, adding surveillance capability to a ubiquitous device while raising privacy questions at scale.