Donald Trump hit out at “woke, Marxist lunacy” in artificial intelligence models, as he ordered his administration to block companies whose systems exhibit “partisan bias or ideological agendas” from doing business with the US government.

This process, known as reinforcement learning from human feedback, is crucial because topics can be ambiguous, disputed, or hard to define cleanly in code. The directives for what counts as sensitive or neutral are decided by the tech companies making the chatbots.