stem/AI/Ethics.md

26 lines
656 B
Markdown
Raw Normal View History

# Fair
- Democracy
- Board-level
- Focus groups
- Committees
- Guidance for best practices
- Consider all stakeholders and impacted social groups
- Within system and the data
# Accountable
- Not moral decision-makers
- Need owners with checks
- Consider impact of system outcomes
- Determine where and when human intervention and approval required
# Private & Secure
- Data management and governance are critical
- GDPR
# Explainable & Transparent
- Protect reputation, build trust
- Demonstrate AI purpose, limitations, risks
- Avoid masking interactions with AI
# Human-Centric & Socially Beneficial
- Social and economic benefit to society