Alignment

In practice, alignment spans the techniques that shape model behavior after pre-training: preference optimization like RLHF, rule-guided methods such as constitutional AI, extensive red-teaming, and safety evaluations before release. Frontier labs formalize the release process in safety frameworks that tie deployment decisions to measured capability thresholds.

As models act more autonomously — as agents operating computers and executing multi-step tasks — alignment work has shifted from filtering bad text toward ensuring trustworthy behavior over long action sequences, where mistakes have real-world consequences.

Commercially, alignment choices are product characteristics: providers differ in refusal rates, tone, and edge-case judgment, and an over-cautious model that declines legitimate requests imposes retry costs and user friction just as surely as an unsafe one imposes risk.