Why we wrote the IICE AI Code of Practice

A few months ago, our team had a quiet but increasingly urgent conversation. The same questions kept coming up from our members. The same uncertainties kept appearing in the wider community. The same gaps kept surfacing in the published guidance from other professional bodies.

How do ethical, force-free canine practitioners use AI?

Not whether they should. Not whether AI is good or bad. Those framings missed the reality our community was actually living with. The practical question was harder and more urgent. Practitioners were already using AI. Some in ways they felt good about. Some in ways they weren't sure about. Some in ways that, on closer inspection, were actually compromising the values that brought them to this work in the first place.

The International Institute for Canine Ethics (IICE) was founded on the principle that the canine sector deserves a professional body that takes welfare, ethics and standards seriously. AI is now a significant part of how canine work gets done. Standards needed to follow.

So we wrote one.

What the Code is

The IICE AI Code of Practice is the first formal code of practice for AI use in canine professional work. It sets out eight principles that protect dogs, clients, practitioners and the wider profession from the harms that poor AI use can cause. It includes a Decision Framework that walks practitioners through the questions they need to ask before using AI for any task. It includes a glossary, FAQs, and a quick-reference card for the moments when a practitioner is uncertain mid-work and needs a fast answer.

The Code is published at canineethics.org and is free to read. We wanted the standard to be available to the whole canine professional community, paying member or not. You don't need to buy anything to access it.

What the Code stands on

Two principles sit at the heart of the Code. We want to share them here, in this blog, so practitioners who aren't ready to read the full document still have the key parts that matter.

Principle 1. The welfare or behavioural judgement on an individual dog is always yours.

AI does not make welfare decisions, behavioural assessments, or clinical judgements about individual dogs. That work is yours. You can use AI to help you organise notes, structure thinking, or check terminology. You cannot use AI to make the assessment. Direct professional observation by you is what makes the judgement professional.

Principle 2. High-risk domains require direct human practice.

Some areas of canine work are too consequential to be supported by AI advice at all. Aggression and bite history. Welfare risk assessments. Veterinary referral decisions. Cases involving children or vulnerable adults. Court reports. Multi-agency cases. Decisions about rehoming or euthanasia. These stay human. No AI.

These two principles are not aspirational. They are the static and integral in practice.

The other six principles in the Code cover critical evaluation against welfare science, intellectual honesty and authorship, transparency with clients and stakeholders, data protection and confidentiality, the continued necessity of genuine professional development, and the ongoing rhythm of ethical reflection. All eight together form a practical, working framework for canine professionals who want their AI use to align with their values.

Why we made it free

The Code is a piece of sector infrastructure. It belongs to the canine professional community, not to the International Institute for Canine Ethics (IICE)  as an organisation.

We expect other professional bodies to follow and develop their own AI guidance. Our hope is that the IICE Code becomes part of the wider conversation, useful to anyone working on these questions, regardless of whether they're a member of IICE or any other body. We will keep iterating the Code as the technology and the sector evolve, and the current version will always be available freely at canineethics.org.

If you're a member of another professional body and you'd like to bring the Code to your community, we'd love to hear from you. Email us at info@canineethics.org. The work is more useful the more places it reaches.

What this means for IICE members and for the wider community

First, the Code is yours to read, share, and reference. We've published it at canineethics.org and you don't need to give us your email address to access it. We've deliberately resisted the temptation to gate it behind a sign-up form. Standards bodies publish standards. They don't trade them for marketing data.

Second, if you'd like a practical companion that walks you through applying the Code in your business, we've also published the IICE AI Playbook. The playbook is available at £9 and is hosted as an Online Programme on the IICE platform. We'll be writing more about the playbook in next week's blog. For now, just know that it exists.

A closing reflection

The most encouraging thing about writing the Code has been the response from practitioners who have read early drafts. Almost without exception, the response has been some version of "thank you for putting this together - it is so important." Ethical practitioners across our sector have been asking these questions privately for some time. The Code makes them shareable, discussable, and accountable.

That's what professional standards do at their best. They take the values we hold individually and turn them into something we can hold together.

Welfare first. Always.

Read the IICE AI Code of Practice