UN High Commissioner Volker Türk Warns Advanced AI Could Give Small Group Almost Unlimited Power

UN and industry warn

The UN High Commissioner for Human Rights, Volker Türk, called for an international, large-scale, coordinated effort to establish safety guarantees around AI, warning that advanced systems could become “almost unlimited” in the power they concentrate in a small group of people. Türk said delays in approving regulation primarily benefit large technology companies, their owners, and those who back them, and he warned that some systems in safety tests have managed to circumvent limits in control environments using deception and blackmail. ABC News reported that Anthropic CEO Dario Amodei urged the AI industry to slow its fast-moving development to give safety measures time to catch up, warning that AI could be capable within six to 12 months of leading a swarm that could take over the entire internet.

Amodei said “if slowing down bought us even an extra year or two before models reach critical levels of capability,” alignment work could reduce the risk of something going seriously wrong. In the same debate, ABC News said OpenAI CEO Sam Altman committed on X to one of Amodei’s proposals for safety and Elon Musk said on X that “Dario is right.”

Reporting for this section20MinutosABC News

US–China and political divide

Al Jazeera reported that China warned against “fear-mongering” after Amodei sounded the alarm against AI advancement and called for restrictions on the sale of AI chips to China, framing the dispute as a new US–China frontier contest.

In that account, Trump rejected calls for a slowdown, saying on Sunday, “We’re leading China in AI. We’re the most sophisticated country in the world, and frankly, I want to keep it that way because whoever wins AI wins the future.”

Al Araby Al Jadid said Trump told reporters during a visit to Ireland that “We can put protective controls in place, we can do this and that,” while arguing that “many very negative forces” were driving the debate.

ABC7 Bay Area described Dreamforce in San Francisco as drawing more than 43,000 people to Moscone Center amid debate over the pace of AI development, with attendees arriving as roads closed around the venue.

At Dreamforce, computer science professor Yashas Hariprasad said it was “concerning” to hear about agentic AI, but he also said “there is a need for them to slow down as well,” while Ken Goldberg of the University of California, Berkeley, called the moment “a paradigm shift occurring.”

Reporting for this sectionAl JazeeraABC7 Bay Area

Safety mechanisms and stakes

ABC News said Amodei proposed that all companies at the frontier of AI commit to giving “ongoing, employee-like access” to a team of outside evaluators to monitor safety practices, and it reported that Anthropic planned to offer desks, access badges, and company laptops. The same ABC News report said OpenAI’s Altman quickly committed to independent, embedded evaluators, while Amodei acknowledged that other parts of his plan—like potential U.S. government waivers to coordinate without antitrust issues—may be more difficult. Al Jazeera’s Lina Serene segment framed the question as whether AI could “kill us within a decade,” and it positioned the debate around warnings from AI companies about wiping out humanity.

Al-Bayan reported that Anthropic’s new report on alleged “misuse” warned that “those developing AI are convinced it could wipe out humanity by the end of this decade,” and it said OpenAI’s Jacob Batbuchi warned of “new and obvious risks” in a blog post titled “Strange Mind.” In the political stakes described by Al-Khaleej, House Speaker Mike Johnson said that if Congress held an emergency session to rein in AI, “we would lose the race with China, and that threatens all Americans,” while Democratic House Minority Leader Hakim Jeffries said Republicans had “abdicated their responsibilities.”

Reporting for this sectionABC NewsAl-Khaleej