Definition
This refers to laws requiring that a human being must always retain meaningful decision-making authority over AI-controlled weapons systems, rather than letting the AI make lethal decisions on its own. A bipartisan US House bill is the first legislative attempt to write this requirement into binding law for the American military.
Why it matters
This sets an early legal precedent for how much autonomy militaries — and by extension, defense contractors and AI suppliers to government — are legally allowed to give AI systems in high-stakes decisions.