Here is today's AI news. Story 1. Meta. Meta launched Muse, a personal AI agent for web, mobile, and WhatsApp that can connect to services including email, calendars, payments, health, shopping, and smart-home tools.. Meta launched Muse, a personal AI agent initially available in the U.S. through iOS, Android, the web and WhatsApp. It can send emails, book travel, fill forms and shop by navigating websites via a dedicated secure virtual machine. Users choose connected services and can revoke access; Stripe Link provides one-time cards for purchases. The free tier has $20 and $100 monthly upgrades, intensifying competition in autonomous consumer AI while raising privacy concerns. Story 2. OpenAI. OpenAI said a system of roughly 10,000 agents produced an analytical proof and Lean formalization related to a forced Navier–Stokes result; the high-profile claim drew calls for independent scrutiny and a dispute over credit.. OpenAI says an internal system coordinated roughly 10,000 AI agents for 88 hours to produce an analytical proof that a forced, three-dimensional Navier–Stokes flow can develop a finite-time singularity while retaining finite energy. Another model formalized the argument in Lean over 17 hours. The claim could reshape automated mathematics, but it remains unverified by independent experts or the Clay Institute. Mathematician Tristan Buckmaster also disputes credit and data-use issues. Story 3. Microsoft Research. Microsoft Research released FrogNano, reporting a 4-billion-parameter software-engineering agent trained entirely through reinforcement learning on synthetic tasks, without distillation from frontier models.. Microsoft Research introduced FrogNano, a 4-billion-parameter coding agent trained purely with reinforcement learning on synthetic software-engineering tasks—without distillation from larger frontier models. Its TaskPilot pipeline generates tasks near the model’s “frontier of learnability” across roughly 1,500 environments. After five 300-task iterations, FrogNano reportedly reached 61.5% on SWE-bench Verified, suggesting smaller, cheaper agents can rival systems many times their size. Story 4. Mistral AI. French AI company Mistral raised a €3 billion Series D at a reported €21 billion valuation, described as the largest equity round in European technology history.. Mistral AI announced a €3 billion Series D on Sept. 8, valuing the French company at over €21 billion—the largest equity raise by a privately held European technology firm, it said. Samsung Electronics led alongside EQT-managed Scaleup Europe Fund and PSG Equity. The funding will expand frontier-model research, computing capacity, infrastructure and international operations, strengthening Europe’s push for sovereign, open-weight AI. Story 5. OpenAI. OpenAI backed mandatory, capability-based national AI-safety rules in the U.S., while endorsing California measures on independent AI assessments, auditor registration, youth protection,. OpenAI urged Congress to enact mandatory, capability-based AI-safety rules—testing, independent assessments, cybersecurity and incident reporting—for advanced models. It also backed California’s SB 813, AB 1405, SB 1119 and AB 1864, covering third-party assessments, auditor registration, youth safeguards and biosecurity. The push could make verifiable safety practices an industry baseline. That's all for today's AI news briefing.