Pharma’s use of artificial intelligence is at an inflection point, demanding a seemingly impossible combination of risk and caution. As you attempt to stay in step with how AI is evolving, you may also need guidance: how much risk is too much? And how can I prioritize safety and compliance while trying something new? Here’s some expert advice on how to get buy-in for AI innovation.
Start with the most effective AI approach
Within3 CEO Lance Hill defined third-generation or combined AI as the power of tools you’re likely already using – like data visualization and dashboards – plus generative AI that improves on those tools by telling you where to go next.
Right now, relatively few pharma companies use this approach. Many are just now implementing second-generation systems that require intensive setup and implementation. Other companies are working to develop their own AI capabilities, but these efforts could take years to reach full efficacy.
Fortunately, the answer to getting started with truly effective AI-supported workflows can leapfrog over both approaches.
“What we’re seeing is that companies who are taking the approach of more of rapid proof of concepts are implementing projects in the third generation,” says Hill. “So they can really get a sense of the level of accuracy and efficacy [that’s possible].”
Prioritize governance and privacy standards
In a panel discussion, AstraZeneca’s Head of Medical Digital Innovation Karsten Friedrich shared advice on the essentials of an AI-focused strategy – including getting leadership buy-in.
“Start small with scale in mind right at the beginning,” said Friedrich. When your pilot project is complete, you can initiate a dialog about AI by presenting a compelling value story. These shared successes foster understanding and support across the organization. Friedrich also recommends keeping governance and privacy standards in mind, as these components will be paramount to the broader organization.
Set realistic expectations
Neeraj Goel Mittal, Global Head of Data, Analytics, and AI at GSK, says that working with leadership requires being clear about timelines and cost-effectiveness. “Sometimes because of the hype [around AI], we do see leadership expects insights to be available,” she says. “And data might not be available in exactly the format we need.”
Mittal also says that the cost-effectiveness piece is a good starting point. “We need to be able to measure the cost-effectiveness…if we can have those metrics, that would be the start of the story. And later on, we can have more metrics to build upon this.”
Make it about patient outcomes
In medical affairs, it may be more effective to tie the impact of AI to patient outcomes. “New patient starts is a metric you can measure,” says Nathan Lear, Director of Commercial Data Science & AI, US Oncology at AstraZeneca. “If I’m using AI to elevate a certain channel and engaging my customer, do I see an improvement in that channel’s effectiveness as it pertains to that measure within medical?”
Above all, start small
AI is the game-changer of all game-changers. But resisting the temptation to change everything at once is key to making successful inroads with peers and leaders.
“If people start thinking – oh, I’m going to bring this huge AI, I’m going to incorporate it everywhere, and insights from everywhere in the world, they’re going to see a very big sum that they need to pay, and then they will just freeze and say – it’s not going to happen,” says Carlos Eid, Executive Medical Director, Cardiovascular, Novartis.
“Start small,” he says. “Choose one therapeutic area. Choose one country. Try it, see if it leads to an impact…and then expand and build on that. Don’t be afraid to start small to just try out.”
Learn how AI is part of a robust insights management strategy: listen to in-depth expert discussions “Can AI reveal real-time insights?” and “The Medical Maestro, Powered by AI” on demand.
