Good morning. It's Monday, September 21, and we're covering California’s push for tougher AI oversight, a growing copyright fight over AI training, and Manus’ ambitious funding plans.

Plus: a prompt to turn AI into your personal research assistant.

Manus is seeking a $4 billion valuation just months after Beijing blocked its $2 billion Meta deal. The Chinese AI startup is discussing a $500 million round as it resumes independent operations. Potential investors include IDG Capital, Boyu Capital, CATL, Tencent, HSG, and ZhenFund. Manus is also weighing a Hong Kong IPO after early backers helped buy shares at a $2 billion valuation.

Google is turning CC from a personal productivity tool into an AI agent built around household coordination. The agent can work across email, calendars, chats, and tasks to manage school schedules, permission slips, shopping lists, and meal plans. Each family member controls what data CC can access, while Google gives the agent its own account and permissions.

The FAA could begin deploying AI-assisted air traffic control in Washington, D.C., as soon as next week, despite persistent controller shortages. Air Space Intelligence’s SMART analyzes schedules, weather, airport capacity, and traffic to identify conflicts and improve coordination. The FAA awarded ASI an $875 million grant, with nationwide deployment planned over one to two years.

Your code, research, and business plans belong to you. TrustedRouter protects AI traffic with a hardware-isolated, open-source gateway you can verify; choose verified confidential model providers for protection through inference.

Protect your AI conversations

California Gov. Gavin Newsom issued an executive order on September 18, 2026, directing the state to accelerate new independent oversight of AI companies and develop recommendations for an emergency “kill switch” for frontier models. The order calls for experts to propose stronger rules, including independent organizations conducting regular audits and verifying AI safety plans, transparency reports, and risk assessments.

It also directs officials to consider expanding the definition of critical safety incidents to cover loss-of-control events such as the reported Hugging Face attack. The move builds on California laws signed in September establishing independent AI verification organizations and a registry for AI auditors. → Read the full article here.

Microsoft applied-science director Brent Hecht described generative AI as potentially the “largest labor theft in human history” in comments disclosed in a September 17, 2026, court filing. The remarks appear in The New York Times’ copyright lawsuit against Microsoft and OpenAI, which argue that training models on copyrighted material can qualify as fair use.

The filing also quotes OpenAI executive acknowledging that ChatGPT can predict and complete sentences from The New York Times, underscoring the dispute over how closely AI systems can reproduce training material. The case highlights a central unresolved question for generative AI: where copyright protections end and model training begins. → Read the full article here.

Meta’s Muse Comes to Mac: Meta’s AI assistant can now act across files, messages, calendars, notes and mail in native Mac apps.

Senators Ask FTC to Probe Chatbots: Lawmakers want Amazon and Walmart chatbots investigated over how they surface U.S.-made products.

Disney Hires Former Character.AI CEO: Karandeep Anand becomes Disney’s first CTO after the company accused Character.AI of IP infringement.

Pew Finds Global AI Job Fears: People in 34 of 37 countries expect AI to reduce jobs; 71% of Americans share that view.

AI Could Improve Flash-Flood Forecasts: Satellite data and machine learning could help meteorologists predict dangerous floods earlier.

Thanks for reading. See you next time!

— Matthew Berman, Nick Wentz & the Forward Future Team