alignment

Ensuring AI behavior matches human values and intentions

“Alignment research focuses on making AI systems safe and beneficial.”

Etymology — From French `aligner` (to line up), from `a-` (to) + `ligne` (line), from Latin `linea`