AI News Roundup – US maintains export ban on Nvidia’s advanced AI chips, new research measures scope of AI-induced “brain rot,” EU mulls changes to AI regulations, and more

To help you stay on top of the latest news, our AI practice group has compiled a roundup of the developments we are following.

• • The Trump administration is maintaining its bar on exports of Nvidia’s most advanced AI chips, according to Reuters. Nvidia’s flagship Blackwell AI chips have been prohibited by the U.S. government from being sold to China, citing national security and economic competition concerns. President Trump had previously suggested that he may permit sales of a scaled-down version of the Blackwell chips, which we slated to be discussed at a summit with Chinese President Xi Jinping last month, though the topic was not ultimately discussed. Nvidia CEO Jensen Huang said during a visit to Taiwan this week that there were “no active discussions” about Blackwell chips being sold in China, following widely-reported comments made by Huang to the Financial Times that “China is going to win the AI race.” Huang clarified those comments, saying that “what I said was that China has very good AI technology,” and that the country was moving “very very fast” when it came to AI. Huang also said that “the United States has to continue to move incredibly fast; otherwise, the world is very competitive, so we have to run fast.”
• • The New York Times reports on new research into how AI may cause “brain rot” in users. “Brain rot” generally refers to low-quality social media content on platforms such as TikTok that are viewed as damaging the mental capacity of viewers, and studies have generally shown a link between social media use and cognitive decline among children. However, concerns are growing that AI usage could have similar effects. A study from the Massachusetts Institute of Technology earlier this year found that, compared to those without AI access, students using ChatGPT to write an essay had lower brain activity and had poorer recall of sentences in the essay. Some researchers recommend altering AI usage in order to reduce such deleterious effects. Another study involving AI search tools showed that AI usage could be less harmful when using it for smaller tasks in a broader research project rather than full reliance on the technology, underscoring the importance of maintaining the activate process of thinking rather than a passive one of accepting AI output alone.
• • The European Union is considering tweaks to its landmark AI Act intended to make it easier for companies to comply with its provisions, according to the Wall Street Journal. Implemented earlier this year (as reported by this AI Roundup), the AI Act was among the world’s first comprehensive AI regulation packages, but faced criticism from industry leaders as hindering AI innovation and harming Europe’s standing in the global AI race. To that end, the European Commission (the executive arm of the EU) will consider several changes to the AI Act later this month, including a grace period for companies with AI products already on the market concerning compliance with an AI output watermarking requirement, as well as clarifications on how the AI Act fits into the EU’s broader digital regulation scheme and a delay in the enforcement of certain fines for noncompliance until August 2027.
• • Bloomberg reports on a new lawsuit brought by Amazon against AI search company Perplexity, alleging computer fraud by Perplexity’s AI agent, Comet, which can make purchases online on a user’s behalf. Amazon had previously sent a cease-and-desist letter to Perplexity, alleging that the company fails to disclose when Comet is shopping for a user in violation of Amazon’s terms of service. The case has the potential to set boundaries on AI agents, which have proliferated greatly in recent years and have gained the capability to perform more complex tasks. Amazon itself is developing its own AI agents, including an AI assistant that can place items into a user’s cart, and AI bots have the capability to disrupt Amazon’s business model as such bots often ignore advertisements. The case is Amazon.com Services LLC v. Perplexity AI Inc., 3:25-cv-09514, U.S. District Court, Northern District of California (San Francisco).
• • A new bill announced this week in the U.S. Senate would require many employers to disclose the impact of AI on their workforces, according to CNBC. The bipartisan legislation, introduced by Senator Mark Warner, a Democrat of Virginia, and Senator Josh Hawley, a Republican of Missouri, would require publicly-traded companies, large private companies, and federal agencies to submit quarterly reports to the federal government detailing hirings, firings, and other workforce changes due to AI. The U.S. Department of Labor would then be responsible for compiling this information into a publicly-available report. Concerns about the effect of AI on the workforce have been growing in recent months, especially as jobs have been cut in many industries, from the tech sector to retail and shipping. Senator Warner said that his bill will “finally give us a clear picture of AI’s impact on the workforce,” and that, “Armed with this information, we can make sure AI drives opportunity instead of leaving workers behind.”