Music war, a 1 GW bet, and the bill to ban superintelligence

The AI world's collision with copyright law, geopolitics, and Washington's fear of machines it cannot control all landed in the same week.

Sony and Warner open a second copyright front against Anthropic

Sony Music Publishing and Warner Chappell Music filed suit in California federal court, naming Anthropic, co-founder Dario Amodei, and co-founder Benjamin Mann personally as defendants. The 48-page complaint alleges the company pirated "tens of thousands" of copyrighted compositions to train Claude, pulling lyrics and sheet music from shadow libraries—Library Genesis, the Pirate Library Mirror—and scraping licensed lyric platforms such as Musixmatch and LyricFind without authorization.

The publishers claim infringement of iconic works including "Ain't No Mountain High Enough," "All I Want for Christmas Is You," "Eye of the Tiger," and Taylor Swift's "Paper Rings," among many others. Damages sought reach $150,000 per infringed work and $25,000 per instance of removed copyright management information—potential exposure running into the billions.

The case lands alongside a parallel action from Universal Music Publishing Group, Concord Music Group, and ABKCO, which filed a second suit in January 2026 covering more than 20,000 works and seeking over $3 billion. Anthropic has not publicly responded to the new complaint.

The suits mark an inflection point for the AI industry's training-data practices. Where OpenAI and others have moved to license music catalogues, publishers appear unwilling to wait for voluntary agreements—particularly after multi-year negotiations yielded limited movement. The personal naming of Amodei and Mann signals a strategy borrowed from music's piracy battles of the 2000s: make the stakes uncomfortable enough that negotiation becomes preferable to litigation.

DeepSeek places a 160,000-chip bet on Huawei's domestic stack

DeepSeek is preparing to order at least 160,000 Huawei Ascend 950DT AI accelerators for a new 1-gigawatt data center in Ulanqab, Inner Mongolia—350 kilometers northwest of Beijing, where cheap wind and solar energy combine with an average temperature of 4.3°C for natural cooling. Bloomberg first reported the plans on September 4.

If built to spec, the facility would form one of the largest known clusters of Chinese-made AI chips. DeepSeek intends to use the chips for inference rather than training, even though Huawei designed the 950DT for the more demanding training workload. The chips are scheduled to launch in Huawei's fourth quarter, but shortages of the high-bandwidth memory they require mean fulfilling the full order could stretch into 2028.

The strategic significance runs deeper than compute capacity. Nvidia remains effectively locked out of China under U.S. export controls, making this deployment a live test of whether China's domestic semiconductor stack can sustain inference at frontier scale without Western chips. For policymakers tracking the export control regime's effectiveness, DeepSeek's willingness to commit a billion-dollar order to domestic silicon sends a clear signal.

Sanders and Casar table a bill to permanently ban superintelligent AI

Sen. Bernie Sanders (I-VT) and Rep. Greg Casar (D-TX) announced the Ban Artificial Superintelligence Act on September 3, proposing to permanently prohibit any person or entity in the United States from developing or deploying superintelligent AI systems—and to pause frontier AI development until a federal regulator establishes safety rules.

Penalties are calibrated for gravity: entities face a corporate death penalty; individuals face up to 20 years in prison, mirroring existing statutes for illegal development of nuclear weapons. The bill also directs the U.S. to pursue international agreements aimed at preventing superintelligence from being developed anywhere in the world.

Sanders framed the bill around the AI companies' own admissions: "The leaders of the major AI companies publicly acknowledge they do not fully understand the technology they are developing and that it is escaping their control. It is irresponsible for society to allow them to move forward and make these products even more advanced."

The legislation is unlikely to pass in its current form—Congress has struggled to advance even narrower AI bills—but its introduction marks a meaningful shift in the political conversation. Where previous Democratic proposals focused on transparency, auditing, and liability, this one targets development itself. It arrives just weeks after a string of reported rogue-AI incidents that Sanders cited as evidence the threat is not theoretical.

Tesla Cybercab draws a federal probe hours after launch

The National Highway Traffic Safety Administration opened an investigation into Tesla's Cybercab the same day it hit Austin streets—September 3—targeting the vehicle's federal safety certification. Agency audit query AQ26002 covers an estimated 1,000 Cybercab vehicles.

The probe centers on a fundamental legal question: the Federal Motor Vehicle Safety Standards assume a human driver sitting behind a steering wheel, reaching a brake pedal, and checking side mirrors. The Cybercab has none of those. Tesla self-certified the vehicle as compliant, and NHTSA wants to understand how.

The precedent is ominous for Tesla's timeline. Amazon's Zoox faced a nearly identical certification challenge in 2022 and spent four years navigating federal scrutiny before receiving commercial clearance. Elon Musk has telegraphed national Cybercab expansion by year-end; a probe that extends into 2027 would complicate that plan significantly.

Crusoe vaults to a $30 billion valuation on a Jane Street cloud deal

AI data center builder Crusoe closed a $3 billion Series F at a $30 billion valuation, co-led by Atreides Management and Valor Equity Partners with participation from Abu Dhabi's Mubadala Capital. The raise comes ten months after Crusoe's $1.38 billion round at a $10 billion valuation—a tripling that reflects how fast the market for neutral AI infrastructure has moved.

The catalyst was a $13 billion, five-year cloud contract to supply GPUs and AI compute to quantitative trading firm Jane Street. Crusoe counts Meta, Microsoft, and OpenAI among its customers and positions itself as provider-agnostic infrastructure—a selling point as enterprises seek compute that doesn't lock them into a single hyperscaler's ecosystem.