Last year I wrote about PROPET from an unusual perspective for me: without a stand, without a product to sell, without commercial objectives. Just with my eyes wide open and wandering the aisles. It was the first time in over twenty years that I attended an industry congress in Spain without being «anchored» to a space, with complete freedom to move around, observe, and talk to whoever I wanted. I left with a pretty clear picture of the sector: active, energetic, but with certain inertias that were already catching my attention back then.
This year I returned to the other side again. With a stand, with KYBERVET, with a real product. But in a different situation from all previous years where I had a stand: almost entirely alone, with no support team, without the possibility of wandering too far or exploring the fair with the freedom I would have liked. Twenty years managing stands with colleagues, with rotations, with the ability to go out and walk the aisles — and this year, tied down. I mention this because it is highly relevant context for what comes next: my view of this edition is geographically more limited than on other occasions, and what I share about the fair in general should be read with that nuance.
That said, what I saw, heard, and experienced these three days — added to what I’ve been accumulating over months of clinic visits, meetings, and conversations with the sector — gives me more than enough material to reflect out loud.
AI was present. But…
I want to be precise here, because the easy impression would be to say that artificial intelligence shone at PROPET 2026. That is not exactly it. It was present, but its presence was scarce given the relevance that, in my opinion, the topic holds today, and what caught my attention the most was not the lack of solutions but the lack of general interest from veterinarians.
What was there? Scribing — automated note-taking during consultations — was probably the most present in the software realm. There was also AI integrated into laboratory analyzers, as a measurement and analysis tool, something that has been consolidating in that segment for a while now. There was an image analysis solution as a complement to X-ray equipment, an application aimed at animal behavior (kudos to those brave colleagues), and a few other specific laboratory tools.
It is not a panorama of silence. It is a panorama of little presence for a topic that, in other sectors, in other markets, in other congresses, already occupies a central space. And what weighs the heaviest is not the number of solutions on display, but the public’s attitude toward them, as they can solve many of the problems they face today, yet I perceive a certain passivity.
On the positive side: there was a pre-congress event that specifically touched on artificial intelligence, and the topic was also discussed in a round table. Two signs that something is moving. Small, but real.
Two profiles that worry me — and they don’t just come from the congress
What I am going to tell you now is not a conclusion drawn from three days at PROPET. It is an observation I have been building over months of visits to clinics, meetings with veterinary teams, and conversations of all kinds. The congress confirms it, but the diagnosis comes from much earlier.
In the sector, there coexist, roughly speaking, two attitudes toward artificial intelligence that worry me for opposite reasons.
The first is the skeptic. The reasons vary: some got burned by a bad first impression, some have a genuine fear that technology will take their jobs, and some simply don’t want to change a way of working that has worked for years. “I do it this way. I’m too old for these things.” I understand and respect it. But there is something this profile isn’t seeing: this time, the change isn’t going to come from within the profession. It is already coming from the outside. The client arrives at the consultation having consulted the AI, with more elaborate reasoning than the old Dr. Google, with more precise questions, with formed opinions. The veterinarian who is not up to par for that conversation not only loses authority — they lose relevance.
The second is the delusional delegator. They want AI to solve everything, even the most complex matters and those requiring the greatest clinical judgment. And here there is a nuance that is frequently forgotten: AI does well what it has been trained to do, and nothing more. Every tool has a scope and limitations. Using it outside that scope is not innovation, it is a mistake. Artificial intelligence can analyze large volumes of data, cross-reference information that no professional would ever have the time to process — from subtle values in bloodwork compared to the patient’s own history, to updated literature on a rare pathology — but it requires a professional who knows the tool, understands its limits, and maintains clinical judgment at the steering wheel, always.
And here lies the paradox that closes the argument: the human brain leans towards cognitive efficiency when given facilities — towards laziness, to put it without euphemisms. Both the one who rejects AI and the one who asks a generalist AI to resolve complex cases it wasn’t trained for are doing exactly the same thing: using AI without supervision, without specialization, without control. One believes they aren’t using AI. The other believes they are using it well. Both are at the same risk level.
And this has a particularly striking dimension in larger centers and corporate groups. The argument for not providing specialized clinical support tools to the team is usually precisely to avoid that dependency — that veterinarians stop reasoning and just do what the machine says. It’s an argument I understand and respect. The problem is that this same control is not applied to the individual use of personal accounts on ChatGPT, Gemini, or Claude, without supervision, without protocol, and without even knowing what level of training each employee has to use them. The result is the opposite of what was sought: instead of protecting the clinical judgment of the team, that judgment is left exposed to generalist tools without validation, without clinical context, and without any kind of internal governance. Providing the right tool and training the team to use it well is, by far, the safest option.
That risk is not abstract — it has real clinical consequences.
Precisely for that reason, training is not optional. It is the difference between using AI as a capability amplifier or as a crutch that, to make matters worse, might break at the worst possible moment.
At the congress stand, the conversations were more balanced than what these two profiles suggest. Open-minded people, willing to listen, interested in understanding. That profile exists. But it is still in the minority.
The value of time — and the mindset that devalues it
There is something underlying all of the above that needs to be stated clearly.
When you talk to a veterinarian about tools that save administrative time, that handle calls that are otherwise missed, that improve the client experience and reduce friction, the reaction is usually one of genuine enthusiasm. “That is exactly what I need.” And then comes the question about price. And the enthusiasm cools down.
This is not exclusive to the veterinary sector, but here it takes on a particular dimension. The profession is systematically undervalued in Spain. In Portugal, the veterinarian is called “doutor”. In France, they have a social and economic recognition that does not exist here. The external causes are well known: too many faculties, an excess of graduates, price competition among clinics that cannibalize each other.
But there is an internal factor that is less talked about and actively contributes to that low recognition: a mindset within the profession itself that does not help. It is hard to invest in progress — in technology, in training, in consulting, in digital tools. Everything “should be cheaper” or outright free. And at the same time, we demand top-tier results and protest when something doesn’t work as it should.
That combination — low investment in one’s own business, high expectations for outcomes — is an equation that doesn’t hold up. Time bears a cost. The missed call has a cost. The client who leaves because you took three days to answer has a cost. The veterinarian with no digital presence loses patients they don’t even know they’re losing.
Until the sector starts calculating those costs honestly, it will continue making investment decisions that seem prudent but are actually very expensive.
Management software: the dilemma of wanting to control everything
Veterinary management software companies are in a position I understand and that until recently was logical, but with which I now fundamentally disagree.
Their logic is understandable: keep the core of the product while adding layers of value with AI functionalities. The problem, in my opinion, is the strategy they are choosing to execute this, and it has two big flaws.
The first is trying to develop clinical functionalities as well, which requires a depth that is not easy to obtain. Training a model for diagnostic support, advanced clinical analysis, or specific pathologies requires huge datasets, real specialization, and years of work. When a generalist software company tries to shortcut that path and the outcome fails — as inevitably happens — the message the veterinarian receives is not “this software is not ready for this”. The message is “AI doesn’t work”. And that damage to perception affects the entire ecosystem.
The second mistake is treating the client as a hostage. Closed systems, blocked integrations, zero freedom to connect third-party solutions… or connections that are only opened by charging significant financial amounts, sometimes bidirectionally (charging the third party and charging their own client). The business logic is understandable — captive clients, guaranteed recurring revenue, and until recently, it could require development that took time and logical costs. But this is a strategy that puts the vendor’s interests ahead of the veterinarian’s.
And here is the problem: the technological barrier to building integrations has dropped drastically. What previously required months of development and large teams is now accessible. Many more specialized solutions will emerge than any generalist software will ever be able to develop internally — animal behavior, dermatology, ophthalmology, oncology, communications management, automated reception. The client is going to want to use them. And if their software does not allow it, the software becomes the problem, the new friction.
The veterinarian shouldn’t have to adapt to the tool. The tool should adapt to the workflow of the veterinarian. That shift in mindset — from an enclosed product to an open platform — is inevitable. The question is who will understand it first.
A side observation: Asian equipment and what it tells us
There was something else at PROPET that has no direct relation to AI but speaks loudly about the mindset of the sector.
This year there was a presence of Asian companies the likes of which I don’t remember seeing in more than two decades of veterinary congresses in Spain, and which I had only witnessed at international events for the purpose of market research and seeking distributors for new players.
This time my impression was different, and I witnessed how several of these companies were selling equipment directly to the end user, cutting out the distributor, with prices at or below the distributor cost limit. Without local support, without presence in Spain, without a guarantee of post-sale service beyond what can be managed at a distance.
And there were queues.
For low-value equipment — one hundred, two hundred, three hundred euros — the equation makes sense. The risk is acceptable and the savings are real. However, seeing genuine interest in investment equipment — five, six, eight thousand euros — paid without an invoice, in cash, and to companies with no local presence or real possibility for claims — that is hard for me to understand.
I am not questioning the quality of Asian equipment; it has improved enormously and there are genuinely good products. What I’m questioning is the complete equation. Every piece of equipment fails; it doesn’t matter where it came from. And when that moment arrives, the initial saving can become a very expensive problem if you don’t have someone backing you up.
The intermediary has a cost, yes. But they also have functions: support, training, warranty, proximity, accountability. When you eliminate them, those functions don’t disappear. They simply cease to be covered.
The congress is no longer a marketplace
A final note that gives context to everything above.
PROPET — like most events in the sector — has ceased to be a transactional event. The shift began to be noticeable before COVID, but the pandemic was the hinge moment that accelerated and consolidated it. Before, congresses were where orders were closed, where distributors launched campaigns, where the most powerful clinics made major purchasing decisions. Real business happened in the aisles.
That is no longer the case. My hypothesis is that the consolidation of the sector into large corporate groups has been the determining factor. The clinics that used to purchase independently at each congress are now part of corporate structures where purchases are centralized, planned in advance, and not decided at a fair. The result is that the congresses have been emptied of transactions and have become relational spaces, for brand building and conversations.
It’s not necessarily a bad thing. But it has major implications for the industry, for those who invest in stands, for the organizers. And it forces us to rethink what can truly be expected from these types of events.
For KYBERVET, PROPET has been visibility, conversations, and taking the pulse of the market. Not direct sales. And with that expectation, it was exactly what it needed to be.
In closing: there is movement, but we need to accelerate it
I don’t want to end on a negative note, because it’s not what I feel.
The AI pre-congress, the round table, the professionals who approached with genuine interest to understand and get trained, and specific projects that are already underway — these are real signs that something is moving. The conversations I held at the stand were of high quality. There are veterinarians who see it, who want to understand it, who are looking for tools that truly help them. That profile exists and is growing.
The problem is not the sector’s ability to adapt. The problem is speed. AI does not wait. The market does not wait. The new generations of pet owners — who demand immediate answers, digital presence, frictionless service — do not wait.
Training is urgent. Honest outreach — neither catastrophist nor naïve — is urgent. And the willingness to invest in tools that genuinely improve real work, the client experience, and the sustainability of the business, is also urgent.
At KYBERVET we have been working precisely on that for months: concrete solutions for specific problems in the veterinary sector. Without empty promises, without magic. With tools that work, that save time, that improve service.
Three days at PROPET — and an entire year of conversations — have confirmed to me that we are on the right path. And that there is still a lot of work to do.
