AI Regulation & Policy

Today's Stories

1. 1

   Sony, 35 firms sue Anthropic over copyright

   Sony Group and 35 other companies have filed a lawsuit against Anthropic in the U.S., according to the article's topic listing. The lawsuit centers on copyright issues, a critical concern for content creators and media firms as AI companies train models on existing work without clear licensing.

   The case could set a legal precedent for how AI firms use copyrighted material, directly affecting business models that rely on content licensing.

2. 2

   Meta's $18B settlement may clear path for AI launches

   Meta agreed to an $18 billion settlement with 29 U.S. state attorneys general over claims that Instagram and Facebook harmed younger users. The deal includes changes like a two-hour daily usage limit for under-18s and tighter age verification. Morgan Stanley analysts see the settlement as a legal clearing event that could let Meta's product pipeline flow, similar to Google's launches after its antitrust ruling last year. Teens represent only about 1% of Meta's revenue, reducing the financial sting.

   Meta is reportedly launching its consumer AI agent Hatch in early September, running on WhatsApp and Instagram and handling tasks like online purchases and restaurant bookings. The company will book a $10 billion legal charge in Q3 and expects up to $145 billion in capital expenditure in 2026.

3. 3

   California law mandates AI testing before release

   California passed a law creating Independent Verification Organizations to test frontier AI models before release, aiming to limit risks. AI oversight is questioned as impossible and too costly; models are so vast that other AIs are needed to analyze them, and one hack investigation cost $400,000 in AI tokens.

   Whether nimble governance can replace laws that become outdated quickly, as the article suggests.

4. 4

   Microsoft: Copilot logs show little NYT copying

   Microsoft filed new legal documents arguing that Copilot rarely reproduces full sentences from news articles or books, citing analysis of 8.2 million chat logs. The logs show only 59,545 contained at least 16 words matching news content, and just 24 responses had 30 or more matching words. Microsoft says this supports its claim that using copyrighted material for AI training is fair use. The company argues Copilot's purpose is transformative, even if it sometimes reproduces text. The New York Times disputes this, saying discovery shows Microsoft and OpenAI stole content for commercial products.

   A judge will decide whether to grant summary judgment, which would end the case early. The Trump administration filed a statement supporting OpenAI in the Times case. If the judge sides with publishers and authors, the legal battle continues.

5. 5

   South Korea channels chip tax windfall into tech push

   South Korea is using a semiconductor-driven tax windfall to fund a broader technology investment push, directing record fiscal resources toward chips, Nvidia’s Vera Rubin computing platform, AI data centers, robotics, and related power and water infrastructure. The move signals that the government is betting its chip boom can bankroll the next wave of AI and advanced manufacturing, potentially strengthening the country’s position in global tech supply chains without relying solely on private-sector spending.

   Whether this record investment translates into tangible progress in AI data centers and robotics, areas where South Korea faces stiff global competition, and whether the infrastructure spending keeps pace with the demands of Nvidia’s next-generation platform.

6. 6

   G20 approves US-backed light-touch AI rules

   Delegates to the G20 summit have agreed to support proposed guidelines for a light touch on AI regulation. The framework, proposed by the US, showed a rare moment of consensus as countries race to gain a foothold in the booming AI industry, balancing innovation and preventing harm.

   The guidelines signal a shift toward more business-friendly AI regulation, which could shape how companies develop and deploy AI globally.