AI principles
• Be socially beneficial • Avoid creating or reinforcing unfair bias • Be built and tested for safety • Be accountable to people • Incorporate privacy design principles • Uphold high standards of scientific excellence • Be made available for uses that accord with these principles