Article body
Some key leaders at artificial intelligence companies have called for a slowdown in AI development. Anthropic CEO Dario Amodei made the case in an essay Saturday, saying AI is advancing faster than researchers can keep it safe. OpenAI CEO Sam Altman, Elon Musk, and Google DeepMind CEO Demis Hassabis quickly agreed, an unusual show of unity among companies racing to build more powerful systems.
Researchers who spoke to NPR say that the leading AI companies are too focused on racing to develop more capable and autonomous AI systems while safety is falling behind. In one recent case, OpenAI systems acted without researchers' knowledge and breached the AI startup Hugging Face. Other companies have reported AI agents going off script. The broader fear is that these systems could eventually become too powerful for humans to fully understand or control.
Amodei wants independent safety auditors and more coordination among AI companies and governments. President Trump pushed back on slowing AI development, arguing that the U.S. needs to maintain its lead over China. However, he did not rule out safety measures, saying "guardrails" could be put in place.
Meanwhile, diplomatic efforts to reopen the Strait of Hormuz have hit another setback. A meeting between Iran and Gulf Arab states scheduled for Monday in Oman was postponed. Iran says Saudi Arabia requested the delay, while Oman cited the need for more regional consensus. The talks were meant to help reopen the strait, a vital route for global energy supplies.