Safety classifier over-blocks a legitimate use case (personal call recording) and benign adjacent topics

Legitimate, legal personal use: recording my own phone calls on my own device (Spain, one-party consent, and the only other party on the call is me). Across a session, the safety classifier repeatedly stopped responses, including on clearly harmless adjacent topics: which commercially-sold phone ships native call recording, and a generic audio-transcription pipeline unrelated to any specific technique. This looks like a false positive: over-broad blocking of a legitimate use case. Please consider recalibrating so legitimate personal-use call recording, and clearly benign adjacent product/how-to questions, are not caught.