San Francisco, September 18 — The fight over how fast artificial intelligence should move has turned inward, and the industry is splitting over it. Divisions have emerged across the tech sector over calls for a coordinated slowdown in AI development, according to an ABC News report. Worries about out-of-control AI and proposals to hit the brakes are not new.
What is new is the volume. Warnings about rogue AI systems causing mayhem have grown louder in recent days, and several leading AI executives have now voiced agreement on the need for a coordinated slowdown — including international collaboration — to protect humanity from worst-case scenarios.
Not everyone is on board. Some tech leaders bristle at the idea that their companies could not ensure safety on their own, the report says.
Amodei’s blueprint
Anthropic CEO Dario Amodei has issued the most detailed plans for slowing development and ushering in coordination across companies and countries. It is a striking position for a man whose company sits among the frontier labs the plan would bind. Anthropic has long positioned itself as the more responsible and safety-minded of the leading AI companies, ever since its founders quit OpenAI to form the startup in 2021.
One of the most widely noted proposals from a recent essay outlining Amodei’s slowdown plan calls for companies at the forefront of AI — often called frontier labs — to commit to giving “ongoing, employee-like access” to independent outside evaluators. Those evaluator teams would get offices, access badges and company laptops so they can monitor safety practices.
Anthropic has committed to doing this. So has OpenAI. Amodei also called for government regulation and intervention, including coordination between frontier AI companies with the help of the U.S. government and the governments of other democratic countries.
His plan extends further, to the U.S. and other countries working to coordinate with “authoritarian governments.” He acknowledged the difficulty, specifically citing challenges in getting cooperation from China. Amodei said he knows accomplishing his proposals would not be easy, but that he believes “we owe it to humanity to try.”
OpenAI’s parallel push
OpenAI has moved on the same terrain. Days before Amodei’s essay was published, the company posted a lengthy text outlining how it was pushing for mandatory national safety requirements and continuing to support state legislation that “strengthens the broader AI safety ecosystem.” OpenAI also expressed support for a coordinated slowdown, though the report does not detail the full extent of that support. The split underneath all of this is simple.
Some leaders agree on the need for a coordinated slowdown and international collaboration. Others resist the implication that their companies cannot ensure safety independently.
The report does not name the executives who pushed back, nor does it pin down the exact timing of the recent statements beyond the sequence that OpenAI’s text preceded Amodei’s essay by days. Watch this space. The proposals on the table — outside evaluators inside the labs, government coordination across democratic capitals, and the far harder question of talking to authoritarian ones — are the first concrete architecture anyone has offered for slowing AI down without stopping it.
Whether rival labs sign on, or keep insisting they can police themselves, is the trend underneath everything now.
