AI in Veterinary Practice: Why Vets Are Paying More Attention Now
Veterinary teams are busy - and the admin load is forecast to increase. The CMA's ongoing market investigation is likely to bring increased requirements around pricing transparency, written prescriptions, and client communication. The practices best placed to handle it are those already using technology to automate the routine work.
The good news is AI is already proving to benefit veterinary teams - 1 in 5 vets are already using AI technology in their daily work, according to BVA's Voice of the Veterinary Profession survey BVA - from diagnostic imaging, lab tests processing to scribes. The next step is providing more tools across the wider team to help in more areas of the business. Your team’s time is finite and valuable and should be spent on clinical care, client relationships, and chargeable work. AI adoption, if done sensibly, has the potential to improve the experience for your team, your clients, and the animals in your care.
Where Is My Data Stored? Understanding AI Data Privacy and Security in Veterinary Practice
Where is data stored?
For most modern veterinary software, data is stored in the cloud - meaning on secure, remote servers rather than a physical computer in your practice. Where those servers are located matters, particularly for UK GDPR compliance. At Lupa, our servers are based in the UK, and we are SOC II certified - an internationally recognised security standard that verifies how a company protects data against unauthorised access, breaches, and misuse.
What happens when AI is involved?
When AI is used to perform specific tasks - such as transcribing a consultation or creating medical summaries - data is processed via external AI providers or our own internal models. At Lupa, we work with a number of providers, including OpenAI, selecting the best tool for each specific AI task.
We operate a zero data retention policy with all of our AI providers. This means they are contractually prohibited from storing or training on your data. It is used to complete the task and no further.
What should you ask your own provider?
Not every AI tool operates to this standard. If you use AI platforms, it is worth asking directly: where are your servers located, are you SOC II certified or equivalent, and do you have a zero data retention policy with your AI providers? Your practice's data is your responsibility under UK GDPR - and a reputable provider should be able to answer these questions clearly and quickly.
For further guidance on data protection obligations for veterinary practices, the ICO (Information Commissioner's Office) publishes practical, free guidance at ico.org.uk.
.png)
What Are AI Hallucinations and Should Vets Be Worried?
AI hallucinations are when AI tools generate information that is plausible-sounding but factually incorrect. A second, less-discussed risk is omission - where relevant information is simply left out. Both matter in a clinical context.No AI provider in the world can guarantee a zero percent hallucination rate - that is a technical reality. What separates responsible AI from irresponsible AI is how seriously this is managed.
At Lupa, we work with leading models from OpenAI, Google, and Anthropic, selecting the highest performing tool for each specific task. Every three months - or sooner if a stronger model is released, we evaluate and update our systems accordingly. Every AI-generated medical summary includes a direct link back to the original consultation, so vets can verify exactly where information has been drawn from. This tool helps extract relevant clinical information - from lab results and diagnoses to active prescriptions and referral letters - giving teams a clear, reviewable snapshot rather than a black box output.
We cannot stress enough that AI in clinical settings should always be reviewed by a qualified professional before being acted upon. With the right safeguards in place, the risk is manageable - and the models are improving rapidly.
Where does AI Actually Help in Veterinary Practice - And Where Should You Be Cautious?
Where AI genuinely helps right now
The strongest use cases in veterinary practice are well-established: clinical documentation, SOAP notes, patient history summaries, diagnostic imaging support, and client communications. These are time-consuming, repetitive tasks where AI performs well and the risk of harm from an error is manageable - provided a clinician reviews the output.
Where to be cautious
AI should inform clinical decisions, not make them. It does not replace professional judgement, and it does not replace the relationship between a vet and their client. Any output - summary, suggestion, or record - should be checked before it is acted upon or filed. That is not a limitation unique to AI. It is simply responsible use of any clinical tool.Used well, AI does the opposite of replacing your role - it frees you to focus on the parts that matter most. Less time on admin means more time with the patient on the table, more meaningful conversations with clients, and more headspace for the clinical thinking that only you can do.
Talking to your team
Adopting anything new takes time, and when a team is already stretched, even a small change to routine can feel like one thing too many. Add to that not feeling consulted, and not understanding how it could genuinely help - it’s easy to see why AI can be met with a lukewarm response.
A good starting point is reminding your team how much they already rely on technology and AI tools they once had to learn. Lab analysers are a good example - most teams would now struggle to imagine a day without them. AI is a continuation of that, not a departure from it.
From there, it helps to be clear that AI is not coming for anyone's job. It cannot replace clinical instinct, professional judgement, or the trust a vet or nurse builds with a client over time.
The most useful question to ask your team before introducing anything is a simple one: what is actually eating your time right now? Where are the frustrations, the bottlenecks, the tasks that feel disproportionate to their value? The answers tend to point directly to where technology could help - and gives the team a stake in the decision from the start.
Lupa's AI team are building AI scribes, invoicing tools to call handling. If you're interested in seeing Lupa's AI features in action, or want to learn more about what our dedicated Veterinary AI Lab is building Book a demo or get in touch.
Keep up to date with Lupa news and special offers.
