PRINCIPLE 01
Augmentation over Automation
AI agents should enhance human capabilities rather than simply replace them - empowering people to achieve more, working alongside humans as partners.
PRINCIPLE 02
Bounded Adaptability over Rigid Programming
Systems that learn and respond to changing contexts, within limits that have been tested - and that hand the decision back to a human when they move beyond them.
PRINCIPLE 03
Transparency over Black-Box Operation
Users should understand what an agent is doing and why - building appropriate trust through visibility.
PRINCIPLE 04
Human-Centred Design over Technical Efficiency
Technical performance matters, but serving human needs, preferences and values matters more. Agents should be intuitive and aligned with human intent.
PRINCIPLE 05
Accountable Collaboration over Isolated Operation
Agents that communicate and coordinate with humans and other agents, while each stays within its own mandate and answerable to people who can see, and stop, any one of them.
PRINCIPLE 06
Responsible Operation over Unrestricted Capabilities
Powerful capabilities must be matched with appropriate safeguards, operating within boundaries that prevent harm while maximising benefit.
PRINCIPLE 07
Paced Improvement over Static Deployment
Agents that learn from experience and evolve through feedback, at a pace we can understand and govern - when capability outpaces assurance, we slow down.
PRINCIPLE 08
Verified Trust over Self-Assessment
Trust in an agent rests on evidence that others can check, not only on the assurances of those who built it - safeguards, evaluations and incidents are open to independent review.
