As global competition in artificial intelligence intensifies, the American company Anthropic is calling for a more cautious approach. In a note published by its research center, the designer of the Claude assistant argues that it should be possible to temporarily slow the development of the most advanced AI systems to allow companies, institutions, and researchers time to adapt to the disruptions they could cause.
This stance contrasts sharply with the dominant strategy in the sector, where major players are increasing their investments to develop ever more efficient models. Faced with this acceleration, Anthropic believes that technological advances could outpace the capacity of states and organizations to effectively regulate these new tools. The company specifically cites the risks associated with a loss of control over increasingly autonomous and powerful systems.
International coordination is needed to avoid a race without brakes.
To address these concerns, Anthropic suggests creating an international coordination mechanism between businesses and governments. The goal would be to allow key industry players to collectively decide on slowing down or temporarily suspending the development of certain technologies when the risks appear too high. According to the company, such a framework would prevent one player from being penalized for taking a cautious approach while its competitors continue to advance unhindered.
Jack Clark, co-founder of Anthropic, compared the current situation to a car with an accelerator pedal but no brakes. He argued that the industry needs tools to act quickly if technological advances outpace the control or understanding capabilities of researchers and public authorities.
A debate that deeply divides the sector
This proposal, however, is likely to face strong opposition, particularly in the United States. Some in the technology industry and among politicians believe that a slowdown could offer a strategic advantage to China, which is also engaged in a global race for artificial intelligence. For these players, the priority remains maintaining a technological edge over competing powers.
Anthropic occupies a unique position within this ecosystem. Founded by former OpenAI members, the company has built its reputation on a discourse centered on the security and ethics of artificial intelligence. This strategy has earned it as much support as criticism. Some American politicians and several competitors regularly accuse it of exaggerating the risks associated with AI in order to promote an alarmist view of the sector.
Regulation still limited in the United States
The US administration is beginning to implement its first regulatory measures. The president Donald Trump recently signed a decree aimed at strengthening the oversight of the most advanced artificial intelligence models for reasons of national security and cybersecurity. However, this measure relies primarily on companies' voluntary participation and does not include binding mechanisms comparable to those mentioned by Anthropic.
At the same time, Washington is also seeking to open international discussions on the safety of artificial intelligence. During a recent trip to China, Donald Trump indicated that he had raised this issue with Chinese authorities. For Anthropic, only global cooperation will prevent economic and geopolitical competition from definitively overriding security imperatives.
Community
Comments
Comments are open, but protected against spam. Initial posts and comments containing links undergo manual review.
Be the first to comment on this article.