GenAI Zürich 2026 was a great reminder that, even for someone working in pharma, the most interesting GenAI lessons often come from outside our own industry.

Novelty hidden among buzzwords

The event at Volkshaus Zürich brought together a multitude of sessions across various industries. One could see offers of established companies as well as innovation brought by startups. It was a great mixture of niche examples, roundtable discussions and interesting conversations about prompt injections over lunch. There was even a hackathon for the bravest genAI adopters.

While I could spot a few novel ideas here and there, some of which I am currently already testing in my daily work, many vendors pitched remarkably similar stories about “enterprise AI agents” wired into the same stack of tools – source code in GitHub, tickets in Jira and workflows in Monday.com.

I think this highlights the main issue with current AI approaches: it is very difficult to stand out. You might have a great solution, but it’s the delivery of it that will get the audience’s attention and interest in the tech aspect. Currently, this might be the noisiest scene to compete in.

New reveals

During one of the sessions I realized: banking as an unexpected mirror for pharma. One of the most eye-opening threads on day one was how aggressively financial institutions are embracing agentic workflows. What struck me was how “trivial” actions – like blocking a credit card can require over 50 manual steps to be taken in the background. Indeed, a perfect scene for an AI solution to shine.

Hearing this as someone from pharma, you can’t help drawing parallels, there is a tone of use cases within the organization and in its external activities, that are heavily regulated but also quite standard (of course with a certain variety here and there). The core pattern  is clear, first its about transforming the process (where possible) and squishing some steps together or even removing them entirely; secondly its automating and orchestrating with policy‑aware agents that keep humans in the loop only where judgement is really needed (at least during the validation and adoption steps).

GenAI Zürich made it clear that other regulated industries are already doing this in production; pharma risks being late if we stay stuck at the level of isolated copilots. Since my work is mainly focused on early development I would be curious to know what is the stand of AI usage in other parts of the pharma companies. Would need to chat with folks at regulatory to see how they use AI there, as it can be the closest intersection with this example.

A thin but important pharma footprint

Pharma and healthcare were not the dominant theme of the conference, but they were present in focused pockets. For sure a few interesting talks about medical devices and a roundtable and that is about it. There was a strong emphasis on regulatory approval, patient safety and data privacy. It was quite different from other conferences that I attended, for example at last BioTechX one would be bombarded by AI drug discovery approaches.

To sum up, what I personally would love to see at the next is a bit more original organisational AI solutions that can address specific pharma bottlenecks. In case some event in this direction is happening any time soon please let us know! At the same time, I hope the organisers keep the strong cross‑industry mix – because it’s precisely by listening to how banks fix credit workflows or how airports think about autonomous systems that we get the best ideas for reshaping our own processes.

Ilya Schneider Chernov