Legal and institutional measures are critical to mitigating the potential negative impacts
of AI.
Personal Information Protection Laws:
– Legality of personal information collection, use, and provision procedures
– Thoroughness of prior consent procedures
Anti-discrimination Laws and Human Rights Regulations:
– Checking for discriminatory elements regarding age, gender, race, religion, etc.
– Identifying ambiguously or subconsciously occurring inequality factors
Other Industry Regulations:
– Verify safety standards and certification procedures by sector (medical, financial,
transportation, etc.)
Practical Evaluation Application: Using Checklists
When evaluating AI systems, it’s necessary to check step by step using guidelines and
legal requirements as checklists. For example, when evaluating a medical AI technology
check,
– If proper legal consent procedures were followed when using patient information
– If interfaces are present to enhance explainability to users
– If there’s a structure that can clearly distinguish responsibility in case of errors
