How AIVAe Is Shaping Our Safer AI Future in Pharmacy

In 2024, we took our first step into the regulatory space by applying to the MHRA’s AI-Airlock programme with a focus on prescription verification. Although we weren’t accepted in that initial round, it marked a defining moment for us: a clear signal that the future of pharmacy lies in not just building AI tools, but in building them responsibly.

Today, we’re proud to have submitted a new application to the AI-Airlock programme, this time with AIVAe at its core.

 

From narrow focus to system-wide support

Our first application tackled a specific pain point: prescription verification. It was a necessary function, rooted in safety and workflow efficiency. But through that process, we realised our vision needed to be broader, more patient-impactful, and more forward-looking.

AIVAe was born from that clarity. It’s an AI-powered assistant for pharmacy teams that supports structured service delivery, SOP adherence, and future-facing tools like patient triage and medicine adherence support. Built using large language models like ChatGPT and DeepSeek, AIVAe is now maturing into a regulated digital product. And that evolution brings with it real regulatory complexity.

 

Why reapply? Why now?

We’re reapplying to the MHRA AI-Airlock because the timing and readiness are right. AIVAe now presents genuine regulatory challenges:

  • It will soon include autonomous prompts that influence care decisions
  • It blends rule-based logic with generative AI and predictive models
  • It may require reclassification from Class I SaMD to Class IIa AIaMD
  • It will soon be deployed in live clinical environments, including NHS pharmacy services

These questions can’t be solved in isolation. They need structured support, regulatory dialogue, and real-world testing. The Airlock provides that rare, essential bridge.

 

What we’ve learned

Applying once taught us humility. Reapplying has taught us direction. We now approach AI not just as a technology challenge, but as a governance responsibility. We’ve built stronger partnerships, clearer risk models, and a sharper focus on post-market surveillance and explainability.

 

What’s next

Whether or not we are accepted, the process of engaging with MHRA is one we deeply value. We will continue developing AIVAe in line with the highest safety, fairness, and transparency standards. And we will continue sharing what we learn.

If you’re working on AI in healthcare and navigating similar challenges, we’d love to connect. Let’s help shape the future of safe, scalable AI in the NHS together.