Derrick Dao warns of licensing

- Policy commentator @derrick_dao warned on X that emerging AI safety licensing regimes could entrench large labs by restricting independent access to models. (x.com) - Dao argued licensing regimes may centralize control over model distribution and approvals, skewing safety policy toward commercial incumbents. (x.com) - That criticism feeds into broader secrecy debates about who decides safety thresholds for powerful models. (x.com)