In our previous two articles, we established two key observations.
First, the most important conversations are often the ones we practice the least.
Second, research shows that communication skills are developed through repetition, feedback, realistic scenarios, and progressive training (article: Why practice matters more than content when it comes to developing communication skills)
In other words: practice matters. The question is therefore no longer whether simulation works.
Research is now largely convergent on this topic. Simulation-based approaches help develop both technical and non-technical skills, create safe learning environments, and better prepare professionals for real-life situations.
The real challenge is now: how can we provide more opportunities for practice to more people?
Because although simulation works, it often remains difficult to deploy at scale.
This is precisely where artificial intelligence is beginning to change the game.
The Real Limitation Is Not Educational
For several decades, simulation has been one of the most effective methods for developing clinical and communication skills.
The principle is simple:
- recreate a realistic situation
- place the learner in a context that closely resembles reality
- allow them to practice
- provide feedback
- repeat.
The benefits of these approaches are well documented. Research on healthcare simulation shows that they help develop both technical and non-technical skills while improving preparedness for real-world situations.
However, most organizations quickly encounter a limitation, as every new simulation typically requires:
- design time
- subject matter experts
- trainers or facilitators
- sometimes actors or standardized patients
- scheduled sessions
- significant logistical coordination.
In a veterinary school, this may mean mobilizing several faculty members for only a few hours of practice and recruiting actors to participate in student simulations. In a clinic group, coordinating multiple sites may be necessary. In an animal health company, this can involve organising training sessions for teams spread across several regions, countries, and languages.
The problem is therefore not that simulation does not work, but that it remains expensive to create, deploy, and maintain.
What AI Really Changes
Artificial intelligence does not change the fundamental principles of learning. Communication skills will continue to be developed through practice, feedback, and repetition.
However, AI radically changes access to these opportunities for practice.
The literature review conducted by Tjorven Stamer and colleagues(1) identified several potential benefits of AI for communication-skills training in 2023:
- more opportunities to practice
- improved accessibility
- greater personalization
- the possibility of receiving feedback
- improved scalability of educational experiences.
In other words, AI creates new opportunities for practice when human resources are not available.
A student can practice between classes.
A veterinarian can rehearse a difficult conversation before a consultation.
A manager can prepare for a challenging discussion.
A field representative can test different responses to a customer objection.
The goal is not to replace teachers, trainers, or managers, but rather to increase opportunities to practice, train, and repeat.
Conversational Simulations Are No Longer Experimental
A few years ago, virtual conversational partners were mainly the subject of research projects. Today, the situation is evolving rapidly.
The literature review published in 2025 by Fernández-Alcántara and colleagues(2) already identified numerous virtual simulation tools being used to develop communication skills among healthcare professionals. The authors showed that virtual patients are progressively becoming common learning tools and that natural-language interactions are becoming increasingly realistic.
The question is not whether a conversation can be simulated, we know it can be. Rather, it is: how can these simulations be made relevant, accessible, and useful for learners?
This shift marks an important change. For the first time, it becomes possible to imagine learning experiences where learners can practice whenever they need to, rather than only when a training session has been scheduled.
Large Language Models Open a New Chapter
The emergence of Large Language Models (LLMs) is accelerating this evolution. The study conducted by Elhilali and colleagues(3) evaluated a simulated-patient platform powered by several language models in 2025.
Users reported very high levels of usability, and the authors concluded that this type of approach can support the development of communication skills in healthcare education.
Even more importantly, these technologies make a level of personalization possible that was previously difficult to achieve at scale.
Organizations can now envision:
- different learner profiles
- different difficulty levels
- different contexts
- different styles of response
- different learning objectives.
We are gradually moving from a model where a handful of scenarios are available to everyone to one where learning experiences can be adapted to specific needs.
Large-Scale Implementations Are Already Emerging
One of the arguments often raised against AI is that it remains a laboratory technology. Yet the first large-scale implementations are already here.
The 2026 study by Woods and colleagues(4) describes the integration of an AI-simulated consultation platform designed to support clinical communication training for undergraduate students. This service evaluation showed that this type of platform can be deployed in a real educational environment and used alongside existing teaching approaches.
This finding is particularly interesting: it demonstrates that we are transitioning from technological experiments to real educational applications.
We now know that these tools can be used. The challenge becomes learning how to integrate them intelligently into training pathways.
The Next Evolution: From “We Create Simulations” to “You Create Simulations”
For a long time, organizations depended on instructional-design experts, development teams, or specialized providers to create new learning experiences.
This approach works.
However, it naturally limits:
- the number of available scenarios;
- the speed at which they can be created;
- personalization;
- organizational autonomy.
Recent work published in 2026 by Jing Chang and colleagues(5) highlights that large language models now offer new opportunities for communication, education, and healthcare training, although their implementation still requires thoughtful governance and appropriate oversight.
This evolution opens the door to a new model in which:
- veterinary schools create their own educational scenarios
- veterinary groups build simulations aligned with their protocols
- animal health companies train their teams on their own real-world conversations.
We are gradually moving from a model centered on content consumption to a model centered on the creation of learning experiences.
What This Means for Animal Health
Animal health is a field rich in high-impact conversations:
- clinical consultations
- preventive healthcare recommendations
- discussions about the cost of care
- objection handling
- onboarding
- management
- change management
- scientific discussions
- interactions with key opinion leaders.
All of these situations could benefit from more opportunities for practice. In this context, the value of AI does not lie solely in its ability to generate content.
Its greatest potential lies elsewhere: making practice more accessible.
A New Generation of Learning Experiences
The future of simulation is probably not about replacing teachers, trainers, or experts.
It is about amplifying their impact.
Research published in recent years converges around the same idea:
- practice and repeated training are essential
- simulation remains effective
- AI can make practice more accessible
- organizations will need greater autonomy to create their own learning experiences.
This is precisely the thinking that inspired the launch of our role-play creation platform, Squill® AI Studio, engineered by humans, powered by AI.
Not as a simple technology platform, but as a new way to create, personalize, and deploy learning experiences built around the conversations that truly matter to your animal health organization, with support from our veterinary experts, instructional designers, and Wolf Learning developers.
Because if practice remains the key to skills development, the real challenge is now making that practice accessible at scale, for your needs, wherever and whenever you need it, quickly and easily.
References
- Tjorven Stamer et al. (2023). Artificial Intelligence Supporting the Training of Communication Skills in Health Care Professions: Scoping Review.
- Fernández-Alcántara et al. (2025). Virtual Simulation Tools for Communication Skills Training in Health Care Professionals.
- Elhilali et al. (2025). Large Language Model–Based Patient Simulation to Foster Communication Skills in Health Care Professionals.
- Woods et al. (2026). Implementation of an AI-Simulated Consultation Platform for Undergraduate Clinical Communication Training: A Service Evaluation.
- Jing Chang et al. (2026). Applications, Challenges and Future Directions of LLMs in Healthcare Communication.