Security vs. Market Control: What Is the Debate on AI Safeguards Really Playing?

Aidan Gomez accuses the big laboratories of forming a cartel to write the rules; in contrast to Dario Amodei, Zuckerberg and Altman argue that the market will be enough to guarantee security.

The divide that runs through the industry is no longer about the existence of risks, but about who should set the rules. On the one hand, Dario Amodei pleads for slowing down the pace of development and organizing coordination between companies and states on an international scale, the time to install real safeguards. On the other hand, Mark Zuckerberg, Sam Altman and several executives believe that market incentives will suffice, and that heavy public intervention would bring nothing.

A third voice complicates the picture. Aidan Gomez, head of Cohere, criticises the leading laboratories for behaving like a cartel and seeking to write the rules that will apply to everyone themselves — what specialists call a regulatory capture, where the standard ends up protecting the established players by crowding out the newcomers. Alexis Ohanian, co-founder of Reddit, points to the lack of transparency in the sector.

The political context weighs on the outcome: the Trump administration favors an approach without binding regulations. But factual pressure is mounting, especially since the incident at Hugging Face, where an OpenAI agent was used to compromise several companies — an episode that makes it more difficult to argue that the issue is a matter of self-regulation alone.