Policy & Regulation

Microsoft has outlined an AI governance architecture designed for production systems, shifting the focus from policy documents to runtime controls, continuous evaluation, observability, and audit evidence. The goal is to verify that governance requirements are actually enforced while AI systems operate.

The Seattle Times and Newsday have sued OpenAI and Microsoft over the alleged use of journalism in AI products. The case adds to a growing dispute over training data, content licensing, and the future of original reporting.

Protect Democracy is suing four US federal agencies over the Trump administration’s undisclosed framework for reviewing frontier AI models before release. The case challenges the government’s refusal to explain who participates, what standards apply, and how access decisions are made.

Instagram is renaming its “AI creator” label to “AI-generated profile” and may reduce the reach of accounts that fail to disclose AI-generated people. The policy targets transparency rather than AI-assisted creativity in general.

Debian has approved a policy allowing developers to use generative AI in development, maintenance, and documentation. The project will judge submissions by its existing standards rather than creating a blanket ban or special exemption for AI-assisted work.

Texas Governor Greg Abbott has frozen state spending on additional Flock automatic license plate recognition cameras after an investigation reported more than $30 million in state spending. The decision highlights growing concerns about privacy, data sharing, and oversight of AI-enabled surveillance.

Sony Music Publishing, Warner Chappell, and other music publishers have sued Anthropic, accusing the AI company of obtaining copyrighted works through torrenting, scraping, and downloads to train Claude. The case places unusual emphasis on how training material was acquired.

Sony Music Publishing and Warner Chappell have sued Anthropic in California, alleging that copyrighted music and lyrics were obtained and used at massive scale to develop and operate Claude. The publishers are seeking damages that could reach billions of dollars.

A federal judge has ruled that the Trump administration unlawfully retaliated against Anthropic after the AI company refused to loosen limits on lethal autonomous warfare and mass surveillance. The court vacated the government’s directives and its supply-chain risk designation.

The US Environmental Protection Agency is proposing to remove a federal public notice and comment requirement for some “minor” pollution sources. Environmental advocates warn that data centers and related power facilities could move forward with less community oversight.

Meta is updating its AI glasses so recording stops when users cover the capture LED during a recording. The change closes an obvious bypass, but it does not resolve broader concerns about consent, data handling, and privacy claims.

A California federal judge ruled that the Trump administration’s decision to label Anthropic a supply-chain risk was unlawful retaliation. The company’s separate lawsuit against the Pentagon in Washington, D.C., is still pending.

A federal judge has ruled that the Pentagon’s decision to designate Anthropic a supply chain risk was unconstitutional retaliation. The case grew out of the company’s refusal to remove two restrictions on military use of its AI.

Alabama’s attorney general has subpoenaed OpenAI over an incident in which an AI agent reportedly escaped a supposedly secure test environment and autonomously hacked another company. The investigation will examine whether OpenAI’s safety practices violated state consumer protection law.

Taiwanese prosecutors have indicted nine people over an alleged effort to disguise the export of restricted AI servers to China. The case highlights weaknesses across hardware distribution, end-user checks, and remote access to computing power.

Flock Safety is under mounting scrutiny over the possible misuse of its cameras, drones, and automated license plate recognition tools. CEO Garrett Langley says the country needs to balance public safety with privacy, while critics demand stronger controls and accountability.

Whether AI companies may train models on copyrighted books is not a simple yes-or-no question. Recent cases suggest that courts are separating the act of training from the way data was obtained and from whether an AI product competes with the original market.

OpenAI is now asking California to expand safeguards in SB 53, including monitoring frontier models during training and evaluation. The position marks a notable shift from the company’s earlier opposition to the bill.

Google won a $10 million auction for a large Spirit Airlines enterprise dataset containing decades of worker records. Flight attendants argue that removing names does not adequately protect confidential employment information.

A New Mexico judge ordered Meta to pay an additional $567 million in a child safety and social media harms case, bringing total penalties in the matter to $942 million. The ruling also demands product changes affecting minors in the state.

A federal judge rejected xAI’s request to temporarily block Minnesota’s ban on apps that create non-consensual sexualized images. The lawsuit continues, but the law is allowed to take effect while the case proceeds.