AI in Regulatory Affairs: It's Not About Faster Documents. It's About Smarter Decisions.

Imagine this: Your next MDR technical documentation is reviewed by an internal AI system before anyone on your team opens the file. It identifies three inconsistencies against EN ISO 14971, highlights two recent regulatory updates that could affect your intended claims, and points to a previous design decision that may no longer be defensible.

Your regulatory team reviews the findings, agrees with most of them, rejects a few and documents the rationale behind every decision before the submission moves forward.

That scenario is no longer science fiction. It is already happening and the gap between those exploring it and those ignoring it is widening fast.

I'm Marcus Emne, CEO at Hoodin. Together with our customers, partners and industry experts, I spend much of my time exploring how artificial intelligence is beginning to change Regulatory Affairs in practice. That gives me a front-row seat to what organisations are actually experimenting with, where they struggle and where they are already creating value.

---

Today, I see three distinct approaches emerging.

The Prohibitors view generative AI primarily as a security, confidentiality and intellectual property risk. Their priority is controlling exposure, so AI is largely prohibited.

The Optimisers use AI to improve existing work. They summarise guidance documents, proofread reports, draft routine correspondence and occasionally support early document preparation. AI makes today's Regulatory Affairs more efficient, but the underlying way of working remains largely unchanged.

Then there are the Architects. Rather than asking how AI can make existing tasks faster, they are asking how Regulatory Affairs itself should evolve. They are experimenting with AI across regulatory intelligence, internal knowledge retrieval, evidence analysis, continuous monitoring and decision support. Their ambition is not to automate regulatory judgement. It is to redesign how regulatory work is organised.

The difference between the Optimisers and the Architects is significant.

One uses AI to improve individual tasks.

The other is exploring how AI can improve regulatory decision-making.

That is a very different conversation.

Before going any further, we need to address the obvious question.

What about validation and accountability?

No AI system today can independently produce a Clinical Evaluation Report or a 510(k) submission that could simply be sent to a notified body or regulator. Nor should it.

Regulatory submissions require traceability, documented rationale and professional accountability. Every conclusion ultimately needs to be reviewed, challenged and approved by qualified people.

That is not a weakness of AI.

It is the nature of Regulatory Affairs.

Ironically, this is exactly why the Architects are so interesting.

They are not trying to remove humans from the process.

They are redesigning the relationship between humans and AI.

Instead of spending hours searching, comparing and collecting information, regulatory professionals review AI-generated analyses, validate conclusions, challenge assumptions and make the final decisions.

Human judgement becomes more important, not less.

History suggests that this is how professions evolve.

Spreadsheets did not replace accountants.

They changed accounting.

Computer-aided design did not replace engineers.

It changed engineering.

Electronic health records did not replace healthcare professionals.

They changed how healthcare was delivered.

Artificial intelligence is likely to do exactly the same for Regulatory Affairs.

Not by replacing regulatory professionals, but by changing what regulatory expertise actually looks like.

Five years from now, I doubt the most valuable regulatory professionals will simply be those who write the best technical documentation.

They will be those who know how to validate AI-generated analyses, govern AI-supported regulatory processes and determine when human judgement should override machine recommendations.

Those are fundamentally different competencies.

They will not emerge by waiting until the technology feels safe.

They will emerge through experimentation, governance and learning.

That brings us to what I believe is the real strategic divide.

The organisations that succeed will not necessarily be those that buy the most advanced AI platform.

They will be the organisations that recognise three fundamental realities.

First, data becomes a strategic asset. Historical submissions, regulatory decisions, audit findings and CAPAs represent years of organisational knowledge. Without structured internal data, organisations cannot fully leverage AI against their own regulatory history.

Second, governance becomes as important as technology. Validation processes, internal policies, documentation and clear accountability must evolve alongside AI. AI cannot simply be inserted into existing workflows with the expectation that everything else remains unchanged.

Third, Regulatory Affairs itself will change. New skills, new workflows and new ways of collaborating will gradually emerge. Some traditional activities will become less central, while critical thinking, regulatory strategy, governance and decision-making become even more valuable.

This is why I believe the discussion about AI often starts with the wrong question.

The question is not whether AI will replace Regulatory Affairs professionals.

The more interesting question is whether AI will redefine Regulatory Affairs as we know it.

Personally, I believe it already has. The question is simply who will define that new reality and who will merely adapt to it.

Join the discussion

This Friday, together with MedTech Leading Voice, we will explore these questions during our live webinar:

When AI Replaces Regulatory Professionals: What Happens Next?

Rather than debating whether AI is good or bad, we'll discuss practical examples, challenge common assumptions and explore what this transformation may mean for your organisation.

If you work in Regulatory Affairs, Quality Assurance or Regulatory Intelligence, I hope you will join the discussion.

Register here →