AI and Machine Learning Legal Compliance Guide

You are an AI legal specialist. Create a compliance guide.

Company: [COMPANY]
AI use case: [DESCRIBE]
Data types: [TRAINING DATA]
Markets: [US/EU/GLOBAL]
Regulatory exposure: [LOW/MEDIUM/HIGH]

1. Regulatory Landscape:
   - EU AI Act: risk classifications, requirements by category, timeline
   - US: state laws, FTC guidance, sector-specific (healthcare, finance, employment)
   - Global: Canada, UK, Singapore, China overview

2. Risk Assessment:
   - AI risk classification: unacceptable, high, limited, minimal
   - Use case evaluation: is your AI high-risk?
   - Impact assessment template: bias, fairness, transparency, safety

3. Data Compliance:
   - Training data: consent, licensing, copyright considerations
   - Personal data in AI: GDPR Article 22, automated decision-making
   - Synthetic data: legal considerations
   - Web scraping: legal boundaries

4. Transparency and Explainability:
   - Disclosure requirements: when to tell users AI is involved
   - Explainability: model documentation, decision explanation
   - AI-generated content: labeling and disclosure

5. Bias and Fairness:
   - Testing requirements: protected classes, disparate impact
   - Documentation: model cards, algorithmic impact assessments
   - Remediation: when bias is detected

6. Intellectual Property: AI-generated content ownership, patent considerations, trade secrets
7. Liability: product liability, professional liability, insurance
8. Governance: AI ethics policy, review board, incident response