this policy does not go nearly far enough to protect us from dangerous and unaccountable AI systems," Patrick Toomey, deputy director of ACLU’s National Security Project, said in a statement.