
If you run an owner-led business or medium enterprise, get our AI policy templates free, with a video on how to tailor them to your business.
You'll walk away with our tried-and-tested AI policy documents and a video showing you what to change so they suit your business. Free.
What changed while your policy sat still
separate governments changed the rules on AI in 2025. Europe, the United States, and Australia.
Dates and sources below
essential practices, introduced here in October 2025, updating the standard your policy was probably written against.
Guidance for AI Adoption, National AI Centre, 21 October 2025
Europe
Obligations for providers of general purpose AI models commenced, along with the code of practice and the training data template.
EU AI Act, in force 1 August 2024, with these obligations applying from 2 August 2025
United States
The 2023 executive order was revoked and replaced with a pro innovation mandate, followed by a national action plan in July.
Executive Order 14148, 20 January 2025, revoking the 2023 order. Executive Order 14179, 23 January 2025. America's AI Action Plan, July 2025.
Australia
The National AI Centre condensed the 2024 voluntary safety standard's ten guardrails into six essential practices for every organisation using AI.
Guidance for AI Adoption, 21 October 2025
The part nobody has written down
Your staff are already pasting company data into AI tools you never approved. Not might be. They are.
This is what we call shadow AI. Your intellectual property, your customer data, your prompts, all of it flowing into systems you do not control.
Nobody approved it because nobody was asked. It came in the way every useful thing comes in, one person at a time, because it made their Tuesday easier.
And when it goes wrong
Every employee is making their own risk decisions on your behalf. When something goes wrong, you do not know where to point.
You already know what you need
In language a board will actually approve.
Because that is the part everyone is asking about.
Before they get in the door, not after.
So the risks are written down somewhere other than your head.
Writing all of that yourself is a project nobody has time for. So it slips. And while it slips, most companies are still running templates written in 2023, against rules that changed in 2025.
Someone is going to ask what your AI policy is. The room will go quiet, and then it will look at you.
The question arrives before the document does. That is the whole problem.
So I have done it for you
Board ready. The document you table at the next meeting instead of the silence.
The part everyone is actually asking about right now, handled separately so it can move faster than the parent policy.
So you can check a tool before it gets in the door, rather than finding out it is already in use. At the back sits an Approved AI Tool Register, so everyone can see what has already been cleared.
Preloaded with 12 common AI risks, so you begin with a full page rather than an empty one. It arrives as an Excel workbook, so it stays a working register rather than a printout.
Your audit trail. The difference between a policy that exists and a policy people have signed.
So your managers know how to implement this and roll it out, rather than forwarding it and hoping.
On top of the six documents, a video where I take you through exactly what to change so it suits your environment. Not a webinar, and not a pitch.
All six are aligned to the six essential practices in the National AI Centre's Guidance for AI Adoption, published 21 October 2025.
What this normally costs
what this pack is worth on its own, and roughly what I am paid to produce it.
GreenCyber engagement pricing
what the large consultancies bill for AI governance work of this kind.
Observed range across competitive proposals
You are getting it free, because the template was never the hard part. The hard part is making it true for your organisation, your vendors, your actual use cases. If you want me in the room for that, reply to the email your download link arrives in.
A client story
An agricultural producer came to me in May 2025 for incident response readiness and a start on ISO 27001. Once we opened the policies, procedures and documentation, much of it was out of date or did not exist. On top of that sat AI exposure, and complex industrial controls and robotics.
There was more to do than anyone expected. I led the work myself, alongside their own staff, whose skills grew with it. They also took on due diligence for an acquisition along the way.
In 14 months, their policies, procedures and documentation were brought up to date. Incident response is fully implemented, and their board has run exercises. They have an enforced AI and agentic AI policy and a third-party risk management framework. They are well underway towards ISO 27001 audit readiness.
Their position is defensible. They are prepared.
"We felt included throughout the whole process. The work delivered a seismic uplift to our cybersecurity resilience and capability, and raised C-level awareness and the board's confidence in our security position."
The agricultural producer
If AI is on your board's agenda, get the Board-Ready AI Policy Pack.
For context
AI risks already written into the register, so nobody has to start from a blank page.
Included in the pack
the year Australia's voluntary AI safety standard came out. A policy written against it is now a version behind.
Voluntary AI Safety Standard, 2024, updated by the Guidance for AI Adoption, 21 October 2025
Take it
The six documents, as a download link on screen and sent to the address you give me.
If you want help making it true for your organisation, tell me when it lands.
Six documents. Free.
Take me to the form
The person who wrote it
Most boards will tell you AI is on the agenda. Almost none of them have a document that says what is allowed.
I have been a CISO for over two decades, led response through more than twenty live incidents, and I hold two national advisory appointments. I write AI governance for boards as paid work, which is the only reason I can give this away and mean it.
Nothing in this pack is theoretical. It is built from AI governance work for a real organisation, de-identified, and extended to the six essential practices.
Take the documents and use them. If the hard part is still hard after that, you know where I am.
Questions
It is free, and there is nothing to buy at the end of it. I give it away because the template was never the hard part. The hard part is making it true for your organisation, your vendors and the way your people actually use AI, and that is the work I am paid for when somebody asks.
Six documents: an AI Acceptable Use Policy, a Generative AI Supplement, a Vendor AI Assessment Template, an AI Risk Register Starter preloaded with 12 common AI risks, a Staff Acknowledgement Form and a Manager Briefing Annex.
Straight away. Your download link appears on screen the moment you submit the form, and the same link goes to your work email, so you have it to come back to.
All six are aligned to the six essential practices in the National AI Centre's Guidance for AI Adoption, published 21 October 2025, which replaced the 2024 voluntary safety standard. Most company AI templates were written in 2023, against a standard that has since been replaced, which is how a policy goes out of date without anyone touching it.
Yes. All six are ready to brand and adopt as your own, so what your board approves reads as your organisation's policy rather than somebody else's.
No, and that is where nearly everyone starts. The aim is not to ban AI, because a ban mostly means people stop telling you what they use. The pack gives you an acceptable use policy, a way to vet a new tool before it gets in the door, and a form your staff sign, so the safe path is written down and agreed.
I did. Over two decades in technology and security, in leadership throughout, response led through more than twenty live incidents, and two national advisory appointments. I write AI governance for boards as paid work, and the pack is built from that work for a real organisation, de-identified.
Reply to the email your download link arrives in and ask me into the room. On its own the pack is worth about $10,000, roughly what I am paid to produce it for a client.
Your name, email, website, company size and, if you give it, your phone number are used to give you the pack, on screen and by email, and to reply if you write back. The headcount band tells me the size of the business the pack is going into, so if you do write back I already know where to start. Your details are seen by the senior team and the service providers who host the site, and they are not sold, traded or passed to partners, brokers or vendors. The site uses the Meta Pixel to measure its advertising, and it does not send what you type into a form.