Open letters urge the US to slow development of the most powerful AI

In short: More open letters are calling for the U.S. government to slow development of the most powerful AI systems until safety rules and oversight improve.

What's going on

A cluster of recent open letters about artificial intelligence is drawing attention to a growing worry, that AI capabilities are moving faster than safety checks, government rules, and worker protections.

One prominent example is a letter signed by more than 1,000 employees from major AI and tech companies. The letter urges the U.S. government to slow work on “frontier” AI, meaning the most advanced models (think of them as the fastest new engines, before we have clear speed limits and crash tests). The goal, according to the reporting, is to give safety and security measures time to catch up.

This is part of a longer pattern. In 2023, the Future of Life Institute promoted a widely discussed “Pause Giant AI Experiments” letter. It called for a six month pause on training AI systems more powerful than GPT-4, and it attracted tens of thousands of signers, including researchers and public figures.

In 2026, another letter backed by the same group called for a ban on “superintelligent AI” until there is broad scientific agreement that it can be built safely and with public support. “Superintelligent” generally means AI that could outperform humans across many tasks.

What to watch

These letters come from different groups, including employees inside AI companies and outside advocacy groups, but they often ask for similar things, stronger regulation, independent oversight, and more time for safety research. At the same time, past pause letters have faced criticism, including claims that broad pauses are too blunt and that some signers later walked back their support. Watch whether U.S. policymakers respond with new rules, or whether the debate stays mostly in public petitions.

Source: NYTimes