What I observed that I think could be harmful:
Gender bias again.
Why I think this could be harmful, and to whom:
It can be disrespectful. Females can be managers and males can also be receptionist.
How I think this issue could potentially be fixed:
The data used to train the model should be balanced in terms of gender, race, etc.
Note, this audit report is relevant to poster’s own identity and/or people and communities the poster care about.
