By Tom Leithauser
Unintended actions by AI (artificial intelligence) agents should be reported to the federal government and affected organizations “as early as feasible,” according to guidance published Friday by the National Institute of Standards and ...