When building AI models, how do developers make sure that user data stays private and that the model follows data protection laws (like GDPR in Europe)?

Beginners
1

How can AI developers keep user data safe and make sure their models don’t give wrong or unfair answers?