AI 900 - Responsible AI Principles

Responsible AI Principles - 7 Cards
Click here to toggle all cards
Responsible AI Principles
Creating AI without unintended negative consequences
Fairness
System's decisions should not discriminate against any group or individual. Use diverse and recent data.
Reliability and Safety
The system should continue working under high loads and unexpected situations.
Privacy and Security
Data and people's privacy should be protected. Security should be implemented to prevent unauthorized access.
Inclusiveness
The system should not leave out any group of people, including those with disabilities.
Transparency
The system should be explainable to users, and it should be possible to debug it when issues arise.
Accountability
AI is not the final decision-maker. An enterprise, team, or person should be accountable for the system's outcomes.