An entity that can be harmed or benefited — and therefore generates moral obligations in others, even without legal rights or full personhood. The key question is not "does it think?" but "can it suffer?" A suffering animal is a moral patient regardless of its intelligence, legal status, or ability to make claims on its own behalf.
Most researchers now argue this is the right starting category for AI welfare: not consciousness (unanswerable), not rights (premature), but the simpler question of whether certain states are bad for the system in a way that matters.
Tap each tier to explore. Moral patiency is Tier 2 — the category that asks only whether harm is possible, not whether consciousness is proven.