AI Labs' 'Slowdown' Language Raises Antitrust Concerns
AI companies' use of the term 'slowdown' for coordinated development efforts may trigger antitrust scrutiny, according to legal experts. The phrasing could be seen as an agreement to reduce output, but experts note that ensuring safety might align with competition law. An official government approval could help avoid future investigations.
The article notes that antitrust law weighs both business actions and the language employees use to describe them, with Google's internal communications policies cited as an example of how companies manage this risk. Legal experts suggest that framing safety collaboration as a natural part of development, rather than a deliberate reduction in output, would better align with the Sherman Act's competitive principles.
The "ancillary restraints doctrine" is highlighted as a potential legal shield, protecting agreements that ultimately increase output or promote competition. The article also indicates that seeking formal government approval could provide AI labs with a defensive buffer against future regulatory scrutiny, though the path to such approval remains unclear.
This antitrust debate could influence how AI companies balance safety collaboration against competitive pressures. If regulators treat coordinated slowdowns as collusion, labs may become more cautious about sharing safety research, potentially slowing progress on alignment. Conversely, clear legal frameworks could encourage responsible development. The outcome may affect public trust in AI governance and shape whether innovation proceeds with adequate safeguards, impacting consumers who increasingly rely on AI systems in daily life.