AI sentiment describes whether assistants talk about your brand in positive, neutral or negative terms when they mention it.
Being named is only half the story. "Brand X is popular but has frequent billing complaints" is a mention that hurts.
Negative framing usually traces back to something an engine read: an old review thread, an outdated comparison, a news story. Finding the source is the first step to correcting it, typically by publishing clear current information and earning fresher coverage.
Sentiment scoring by machine is imprecise, so reading a sample of the actual answers remains the best check.