At the Veeva R&D and Quality Summit in Copenhagen this May, Olivia Friett spoke with Crystal Allard, senior director of government strategy at Veeva Systems, about the growing role of AI in life sciences and the regulatory considerations shaping its adoption.
Veeva
There's AI on the sponsor's side and AI on the regulator's side evaluating each other's output. When both parties rely on similar tools, where does accountability lie if something goes wrong? And how do you guard against teams placing too much trust in AI rather than the underlying evidence?
Our customers understand that they hold all of the responsibility for submitting correct data to a regulatory agency. As far as Veeva goes right now, we are making sure that we work very closely with them. We have to clearly define the way that our AI will be working as we roll it out, whether or not it is validatable, and how we'll maintain compliance with those requirements.
We take a careful review of those to ensure that we're remaining compliant as we're building, and we have plans to reach out to both FDA and EMA and ensure that if we have any questions about whether or not our systems are meeting those requirements, that we get answers from them. FDA is fully responsible for making a determination of safety and efficacy on an application that they're reviewing, they are exploring AI tools.
The FDA has undergone significant changes in recent years, both in terms of personnel and policy direction. Is the situation different in Europe, where the regulatory landscape appears more stable?
The EMA has stayed their path where FDA has curved a bit recently.
FDA has been very open about what they're trying to do with AI. It's not as easy to predict where it will land and so they're giving more information. EMA is not putting out a lot of information about how they're using AI. We know they are, but they're not openly telling us what they're implementing and how they're using it.
FDA openly said they're going to use it for filing, so yes, there are major differences in the amount of transparency.
HL7 has created duplicative standards, one for FDA and one for EMA, and I would like them to stop doing that. This over complexity makes it harder for us to build technology that's joyous to use. We're also working with standards organisations to ensure that as they're building new standards, they're assuming AI.
I don't think we're ever going to get away from standards entirely.
We should be assuming that the future of data exchange will be digital and will look different and won't be this over-the-fence process where we do everything here and we validate it, and then we send it over there and then they validate it too.