OpenAI and Anthropic Back Petition Urging US to Slow AI Development
Two giants in the artificial intelligence (AI) race, OpenAI and Anthropic, have formally given their backing to a petition asking the United States government to intervene and slow the development of AI technology if its progress is deemed too rapid and risky.
The petition, released on Tuesday (28 July) local time, was signed by hundreds of employees from both companies and also involved staff from Google, Meta and other technology industry leaders. The move marks a rare moment in which fiercely competing companies agree on the need for stricter external oversight.
The open letter warns that the industry is already very close to the stage of automating AI research. This development is feared could allow the technology to advance exponentially beyond human control.
“We ask the US government to support international efforts to develop the technical and governance tools needed to deliberately regulate the pace of frontier AI development,” the petition states.
International coordination efforts to limit AI development are regarded as an enormous diplomatic undertaking. Some experts liken it to the implementation of nuclear arms control treaties, but in the form of software that can be freely distributed via the internet.
Although OpenAI CEO Sam Altman did not personally sign the petition, he said on a podcast that regulating the pace of AI development may indeed be necessary. OpenAI is targeting the construction of an automated AI researcher by early 2028.
Not everyone in Silicon Valley agrees with slowing down, however. Many argue that AI technology must continue to be pursued aggressively in order to discover new cures for diseases, drive innovation, and preserve America’s military edge in the geopolitical rivalry with China.
Andrew Yoon of CivAI noted that the involvement of big names on the list shows that AI laboratories and their staff are now explicitly beginning to discuss international governance as an urgent necessity before AI models reach the point at which they can no longer be safely controlled.