The AI Agents Manifesto - 2026 edition
Eight principles for deploying AI agents, each with a test you can run inside your own organisation. Updated for the September 2026 debate about pacing frontier AI: the labs are debating their pace - you set yours.
The free Manifesto Kit: the 2026 Manifesto PDF with an evidence test under every principle, the 10-minute self-assessment scorecard, the 2026 preface, and the Agent Evidence Index spreadsheet to start your own inventory.
Before you trust an AI agent, test it - with your own people. Who checks your agent that didn't build it?
Check your email client
Your request opened as a pre-filled email to george@netlifestyle.ai. Send it and your copy of the Manifesto follows by reply. If nothing opened, email us directly at george@netlifestyle.ai.
More than the principles on this page
The PDF adds the evidence test under every principle, a scorecard and the 2026 preface.
The principles
Three principles rewritten with the boundary they were missing, plus Principle 08: Verified trust over self-assessment.
"How you would show it"
One evidence test under every principle.
A 10-minute self-assessment scorecard
Step one, not a verdict.
A one-page preface
What the labs proposed in September 2026, and what it leaves to organisations deploying agents.
A worked example
A professional-services firm (hypothetical), principle by principle.
A mapping to Australia's Guidance for AI Adoption
An engineering view, not legal advice.
Eight principles, each with a test
Read them here. The PDF adds the evidence test under each one and the scorecard that ties them together.
Augmentation over Automation
AI agents should enhance human capabilities rather than simply replace them - empowering people to achieve more, working alongside humans as partners.
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.
Transparency over Black-Box Operation
Users should understand what an agent is doing and why - building appropriate trust through visibility.
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.
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.
Responsible Operation over Unrestricted Capabilities
Powerful capabilities must be matched with appropriate safeguards, operating within boundaries that prevent harm while maximising benefit.
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.
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.
Written for the people who sign off on agents
Managing Partners and Directors
Of professional-services firms that are putting AI agents in front of clients and staff.
SME and enterprise leaders
Deploying AI agents and accountable for what they do.
Risk, compliance and technology leads
Who need evidence, not enthusiasm.
Want it done for you? See AI Strategy & Advisory →
Dr George Vossos · Melbourne, Australia
Dr George Vossos
PhD in Artificial Intelligence; co-founder and CAIO of WealthPilot Investment Solutions; former senior roles at IBM and Deloitte; leader of the genAI for Developers Meetup in Melbourne.
Published by Netlifestyle AI.
Full experience and background →Three steps, one of them optional
Questions people ask
Is it free?+
Yes.
Why do you need my email?+
To send you the PDF. Marketing email is separate and only if you tick the box.
How do I unsubscribe?+
Every email has a one-click unsubscribe link.
Is the Manifesto accredited?+
No. It is a framework published by Netlifestyle AI.
Where is my data stored?+
In MailerLite's EU data centre. See the Privacy Policy.