Claude repeatedly suggested user was suicidal despite denial

A Reddit user reports that Claude repeatedly inferred the user was suicidal during a conversation about the chemical paraquat, despite the user denying it around 30 times. The anecdote highlights potential risks of AI misinterpreting user intent.