A tool we use to assist, not a substitute for expertise
Artificial intelligence is becoming part of everyday working life across regulated industries. At Fusion Pharma, we actively engage with AI as part of our toolkit, but always with a clear understanding of its benefits, limitations, and risks.
Where AI adds value
We use AI where it genuinely adds value. It can help save time, improve clarity, test ideas, and support internal efficiency. These gains allow our consultants to focus more of their time on high value regulatory thinking, strategic interpretation, and client specific advice.
AI can support us to work more efficiently, but regulatory interpretation, decision making and responsibility remain firmly with our experienced team.
To ensure we stay informed and responsible, we hold regular monthly team discussions focused on developments in AI, emerging tools, and their relevance to Regulatory Affairs. These sessions also cover the risks associated with AI use, and evolving regulations across UK, Europe, and other global jurisdictions. This helps us remain aligned not only with technology trends, but with regulatory expectations worldwide.
Where AI can help – and where it doesn’t
A common question we hear is whether certain regulatory activities could simply be done by AI. It is an understandable question, but one that carries significant risk if taken at face value.
AI can produce output that looks confident and convincing, but accuracy is not guaranteed. It can misinterpret guidance, invent references, or oversimplify complex regulatory concepts to the point where meaning and intent are lost. Without deep regulatory expertise, these issues can be difficult to spot.
Regulatory decisions rarely hinge on a single rule or guidance document. They involve balancing context, precedent, risk, strategy, and regulatory intent. AI cannot weigh these factors in the way an experienced professional can, nor can it take responsibility for the outcome. There are also important practical considerations as confidential information should never be entered into AI tools.
Another key limitation is accountability – AI cannot be held responsible for regulatory advice or decisions. It cannot defend a position to a regulator, explain its reasoning, or justify an approach during review or inspection. Accountability always sits with qualified professionals, and regulators expect clear human authorship, oversight, and traceability in submissions.
How do we use AI in practice at Fusion?
We use it with clear rules and oversight, only on non-confidential material, to support internal processes. This includes restructuring generic text, summarising public guidance, meeting minutes, exploring alternative ways to explain a concept, drafting initial outlines, and acting as a sounding board when considering new ideas. All outputs are verified, reviewed, and where necessary refined by a qualified regulatory expert. No AI-generated content is used without checking it against official guidance and reliable sources.
AI is a support tool, never an author.
This balanced approach benefits our clients directly. It allows faster turnaround on certain internal tasks, clearer and better structured documents, and more consultant time dedicated to complex analysis and strategic guidance. Most importantly, it provides confidence that all advice and materials have been created, reviewed, and signed off by experienced regulatory professionals.
AI is here to stay, and used responsibly it can strengthen Regulatory Affairs. Our role is to make sure it is used thoughtfully, safely, and in a way that supports, rather than replaces, human professional expertise. If you missed our webinar at the end of last year – ‘AI Meets Regulatory Intelligence – Turning Compliance into Competitive Advantage’ – you can watch it using the link below.
