Artificial intelligence (AI) is becoming a routine part of veterinary practice. From writing medical records to communicating with clients, the technology is helping veterinary teams handle a growing range of tasks. However, a new survey suggests many practices are adopting AI faster than they are developing the processes needed to use it effectively.
The 2026 State of AI in Veterinary Medicine report, conducted by Digitail and administered by the American Animal Hospital Association, found that 83.7% of veterinary professionals surveyed had used AI tools or software at work, up from 39.2% in 2024. The online survey collected responses from 1,730 veterinary professionals in the United States and Canada between July 27 and Aug. 12, 2026.
The findings point to four practical steps veterinary practices can take to manage AI use, protect information and make the technology more useful in daily work.
1. Put Someone in Charge of AI
As AI becomes more common, practices need a plan for how it will be used. Yet among respondents who had insight into practice-level decisions, 47% said their practice had no AI plan.
The survey also found gaps in basic support for AI adoption. Only 28% of respondents reported having a designated person responsible for AI adoption. Nearly one-third said their practice lacked all four supports assessed: staff training, written usage guidelines, someone responsible for AI and a formal process for evaluating tools.
Practices don’t necessarily need to create a new position to address this gap. Assigning an existing team member to oversee AI use can help ensure someone is responsible for evaluating tools, answering staff questions and reviewing how well the technology is working.
That person can also help the practice establish consistent expectations. Without clear direction, individual staff members may choose different tools or develop their own approaches to using them, leaving the practice without a shared process.
2. Set Clear Guidelines and Keep Human Oversight
Accuracy and data security remain major concerns for veterinary professionals. In the survey, 67% identified AI accuracy and reliability as a concern, while 59% cited data security and privacy. Another 47% were concerned about a lack of training or knowledge.
Clear guidelines can help address these concerns. Practices should establish what information staff can enter into external AI tools, what information must remain within approved systems and when AI-generated content needs to be reviewed before use.
Human oversight is particularly important when AI contributes to medical records, client communications or clinical decision-making. AI-generated content may need correction, and staff should understand that responsibility for checking the output remains with the people using it.
Written guidelines also give staff a consistent reference point as new tools become available. Rather than leaving everyone to decide independently what is appropriate, practices can establish expectations that apply across the team.
3. Choose AI Tools Carefully
Not all AI tools offer the same capabilities, safeguards or compatibility with existing practice systems. The survey suggests some practices could be more systematic when deciding which tools to use.
Only 21% of respondents said their practice formally evaluated AI tools, including data handling and security, and reviewed their performance after adoption. Another 16% said they adopted tools with little comparison, often based on price or what was bundled with existing software.
Before choosing a tool, practices should consider how well it fits their needs, how it handles data, whether it works with existing systems, what support the vendor provides and what it costs. They should also review performance after implementation rather than assuming a tool will continue to meet expectations.
For diagnostic applications, veterinary-specific validation deserves particular attention. Practices should look for evidence that a tool has been evaluated for its intended use rather than relying solely on vendor claims or general demonstrations of AI capability.
A consistent evaluation process can help practices avoid adopting tools that introduce new risks or create additional work for staff.
4. Improve Workflows and Train the Whole Team
Even a useful AI tool can create frustration if it doesn’t fit into everyday work. The survey found that about 65% of respondents said their practice introduced AI without changing workflows or made only small adjustments. Just 10% said their practice regularly redesigned workflows when adopting new technology.
The challenges show up in routine tasks. Among AI users, 56% said reviewing and correcting AI-generated output was a source of friction. Another 44% struggled to transfer information between systems, while more than one-third reported inconsistent outputs or the hassle of switching between tools and logins.
Integration with practice information management systems remains uneven. Overall, 57% of respondents said they still moved some information manually between systems.
These extra steps can eat into the time AI is supposed to save. Staff may need to check a generated note, copy information into the medical record or correct an inconsistent response. Practices should look at how information moves through their systems and identify where better integration or changes to existing processes could reduce duplicate work.
Training is just as important. AI use has expanded beyond veterinarians to practice managers, veterinary technicians, assistants and receptionists. Among receptionists and customer service representatives, reported use rose from 28% in 2024 to 80% in 2026.
Training should reflect that broader use. Administrative staff need guidance on AI tools used for client communications and scheduling, just as clinical staff need to understand the appropriate uses and limitations of tools that support documentation or clinical work.
A More Deliberate Approach May Deliver Better Results
The survey also examined AI maturity, or how well a practice has put the pieces in place to use AI effectively. This includes having a plan, training staff, choosing tools carefully, setting guidelines and fitting AI into everyday workflows.
Practices using AI scored an average of 2.67 out of 4 on the maturity index. The report found an association between maturity and reported business outcomes: practices at the highest maturity level reported an average of 4.05 positive outcomes, compared with 0.64 among those at the lowest level. Nearly 71% of the least mature practices reported no measurable impact from AI, compared with about 12% of the most mature.
The most commonly reported benefits included more time for patient care or higher throughput (44%), improved client retention or satisfaction (32%) and expanded digital services (29%).
How a practice introduces and manages AI may be just as important as which tools it adopts. However, the survey relied on self-reported responses collected at one point in time, so it cannot establish that a more organized approach directly caused better outcomes.
AI use is already widespread among the veterinary professionals surveyed. The next step is making the technology work well in everyday practice. Clear responsibility, practical guidelines, careful tool selection and better workflows can help veterinary teams use AI more consistently while keeping human oversight in place.
