Could GenAI take over regulatory dossier creation?

Renato Rjavec, ArisGlobal, discusses how generative AI (GenAI) technology is transforming everyday regulatory and safety processes, thanks to its ability to summarise key insights and findings from across huge, diverse bodies of content and data.

Pilot solutions can now help companies pre-empt Agency queries and build stronger marketing authorisation applications at speed, drawing on previous health authority interactions. Across 23 different languages, more than a dozen fields of data have been extracted with 90% accuracy – with up to 80% faster processing and three times fewer handovers.  

The technology is also demonstrating powerful potential in monitoring and applying the latest market intelligence, e.g. as part of regulatory impact assessment/change management. Initial applications here too have yielded 50- 80% faster processing, and in this case half the handovers compared to manual regulatory intelligence lookup and response. 

But GenAI’s greatest scope is still to come – within the next couple of years. This will be the point at which regulatory teams are able to lean on the technology to collate and cross-check entire regulatory submissions automatically, with a quality review from Regulatory professionals requiring just a fraction of the effort expended today.  

Large pharma organisations can generate some 600-800 submissions each month. Even a very modest time saving, of just 1-2 hours per submission, would make a substantial difference to associated cost and resource allocation. (Despite the increasing trend of data-oriented submissions, the reality of content-based dossiers is here to stay for some considerable time.) 

Getting from here to there 

Interim use cases will include the provision of inline regulatory guidance to help users in submission compilation; generation of new draft submission content based on existing content; and cross-validation of final content against regulatory guidance and data. Via instant cross-referencing (e.g. with regulatory information management (RIM) system, to ensure that the correct excipient/ingredient information has been used), GenAI will not only expedite submissions compilation; it will also improve the quality and success rate of submission updates. 

Advanced automation could transform the efficiency of dealing with less developed markets, too. Emerging markets, many of which rely heavily on non-electronic files and lack format standardisation, together account for a sizeable proportion of the global life sciences opportunity. Growth in pharma sales in emerging markets is set to accelerate over the next decade, with medicine use in Latin America and Asia expected to rise faster than other regions over the next five years.  

The ability to streamline associated submissions with advanced end-to-end automation promises to be very powerful across these markets, to help companies navigate the differing requirements, deduce “what good looks like”, and swiftly collate and format what’s needed.  

Laying solid foundations now 

Other opportunities for GenAI in a Regulatory Affairs context include automated cross-checks to identify discrepancies and anomalies in data and its formatting, as part of companies’ efforts to get their IDMP data standardisation in order. Further possibilities include more efficient and effective maintenance of labelling compliance internationally across the product lifecycle, again boosted by automated, GenAI-enabled cross-referencing.  

Testing out the possibilities will give companies a feel for how far GenAI can go, how quickly results can be reliably honed, and how much time and budget this could buy back for hard-pressed Regulatory teams. Merely adding a GenAI capability isn’t all it takes, though. Companies will need to bolster their regulatory intelligence knowledge bases (e.g. non-public information and soft intelligence that has accumulated within companies based on their experience and direct HA relations). They will also need to continue the work they are doing to clean up, standardise, and unify their product data. The better the assets GenAI can draw from, the more reliable and transformational associated process automation initiatives will be.