OpenAI quietly removed a ban on "military and warfare" use from its usage policy, as part of a major rewrite to make the policy "clearer" and "more readable" (Sam Biddle/The Intercept)