Anthropic to require ID checks for Claude; Apple's AI-filled iOS 27

Starting July 8, Anthropic will ask Claude Free, Pro, and Max users to verify their identity with a government photo ID and a live selfie, handled by the vendor Persona — making Anthropic the first major AI chatbot company to formally collect biometric data at the consumer tier. The company says the checks are to prevent abuse and meet legal obligations; analysts note they could also create a verified-US-citizens-only path to restore Fable 5 after the government's export-control order. User pushback has been immediate.

Apple is weaving AI through iOS 27 rather than concentrating it in Siri, with the features now live in the developer beta ahead of a public release this fall. Highlights include splitting restaurant bills from a photo of the receipt, automatically upgrading weak or breached passwords, one-tap reminders and calendar events suggested inside Messages, and Safari grouping your tabs by topic. Apple says the processing stays on-device for privacy.

Japan's Sakana AI launched Fugu, a single API that learns to assemble and coordinate a pool of specialist models on the fly instead of making you choose one. Sakana says its standard and Ultra tiers rival frontier systems like GPT-5.5, Gemini 3.1 Pro, and Claude Opus 4.8 — scoring 73.7 on SWE-bench Pro and 93.6 on a long-context recall test — while routing each task to the cheapest capable model. It is grounded in two peer-reviewed papers but ships as a commercial product, though not yet in the European Union.

Switzerland's public AI initiative — a partnership of EPFL, ETH Zurich, and the national supercomputing center — released Apertus Mini, a family of 16 small open models from 500 million up to 4 billion parameters, built by distilling its larger 8-billion-parameter model. Everything is open: weights, training data, code, and methods, with support for more than 1,000 languages and a design that meets European Union AI Act rules. The small sizes are meant to run on phones and embedded devices rather than in data centers — a sovereignty-minded counterweight to closed American and Chinese models.

A widely upvoted bug report says OpenAI's Codex coding assistant writes enormous volumes of low-value logs to a local database — one developer measured about 37 terabytes in 21 days, which works out to roughly 640 terabytes a year, enough to wear out a typical consumer solid-state drive inside a year. The cause is a verbose logging default that captures internal traffic most users never need. Related complaints stretch back across several versions of the Desktop and command-line apps, and the issue remains open and unfixed.