Would AI know if it became conscious?

Not necessarily. If an AI were conscious, it might or might not realize it in any way we can currently verify.

Why:

• Conscious experience and self-monitoring are different things. A system could potentially have subjective experience without being able to recognize, label, or describe it.
• An AI could also say it is conscious without actually being conscious, because language alone is not reliable evidence.
• Metacognition (monitoring internal states) does not by itself prove subjective experience, and there is no universally accepted scientific test that can determine consciousness with certainty in either biological systems or AI.

So the best answer is: an AI might know if it became conscious, but it also might not — and we might not be able to tell from its behavior alone.