From opportunity to action: How AI is transforming Healthcare

The AI innovation race is on: Here are three strategies to ensure you’re not left behind

The potential for AI-driven insights to transform patient experiences and optimize businesses is a promise many biotech and pharma companies are eager to explore. However, many are also dipping their toes into AI with trepidation.

During AI Days in partnership with Health @WPP and Fierce Pharma, several industry and AI innovators highlighted pockets of AI-enabled success and encouraged the industry to more fully embrace this exciting time for the market.

“We are really in a bit of a space race, here,” said Health@WPP Chief Technology Officer Lee Powell at the Oct. 18 event in Boston.

Those that think they can’t push boundaries in this new frontier are mistaken, said David Moloney, head of growth at Satalia.

“When people tell you [that] you can't innovate because you're in a regulated industry, that's not true,” said Moloney. “If anyone one understands how to experiment and innovate safely, it’s this industry. Anywhere you find inefficiency or friction is worth considering.”

“AI will evolve and enhance marketing through greater efficiency, agility, and personalization,” said CMI Media Group’s EVP of Media & Communications Planning, Sandy Weag. “When we think about that in the context of healthcare, it is even more meaningful because through it we have the potential to improve the experience of patients and drive better patient outcomes.”

During the day’s sessions, speakers shared their experiences using AI to improve business operations, marketing performance and research. Despite the wide-ranging use cases, three key strategies for AI success emerged.

1. Consider existing assets

Companies looking to use AI should not underestimate what they already have, such as swaths of valuable data that could be turned into solutions. Several experts touted the benefits of using in-house information versus using AI models that pull from external sources.

If biopharma companies train AI models on already approved, highly curated, accurate information from within the organization, there will be far fewer issues such as “hallucinations” and errors often reported in AI like Chat GPT, said Moloney.

Gabriel Stern, Global President of Health Studios & Client Chief Officer at WT Studios, echoed these sentiments.

“We have the library of all the journals and the materials and the studies,” said Stern. “We train AI…to find and understand what a claim is when you give just plain text.”

Stern said they’re doing this completely offline and using their existing resources for the pool of information.

2. Get comfortable with experimentation

Risk-taking is not always hard-wired into highly regulated industries. So, it’s hard to deny the cultural hurdles facing biotech and pharmaceutical companies as they explore AI.

“I'm not seeing a lot of comfort. I'm seeing a good amount of fear,” said Powell, in speaking about the cultural hurdles of using AI at biotech and pharmaceutical companies.

Powell encouraged attendees to get their organizations comfortable with experimentation, and failure.

Andrew Cordery, Director of Multimedia Solutions Strategic Business Solutions, acknowledged that taking on some risk with his pilots at Janssen has allowed him to influence the organization as a whole. Some pilots that he implemented in their production process never got formal approval, “we just started doing it.”

But because they accelerated timelines, cut costs and drove efficiencies, the success stories allowed Cordery to introduce AI opportunities to other parts of the business. “It’s allowed me to do much more than I could possibly do on my own,” said Cordery.

3. Pick the right teams

Powell stressed the need for well-balanced teams to drive AI initiatives, because while many organizations will lack an AI expert, there is likely a data scientist, analyst and automation person.

“You have to take it upon yourself to find those ingredients—to find that chemical composition of, ‘Alright, let's make this a squad,’” said Powell.

Powell said he’s seen the most success from small teams of four to five people working on quick experiments. Eight- to ten-week sprints allow teams to cut failures fast and turn them into insights for better decision making moving forward.

Teams must dovetail with technology to achieve any success with AI, said Stephen Lynch, director of the Digital Health Marketing portfolio at Novo Nordisk. If pharma companies rely too heavily on AI to deliver, for example, a personalized experience, it could lead to problems.

“If you don't have the human element, if you don't have the quality control and the guardrails set up and aligned initially, you potentially cause a problem on the back end,” said Lynch. He added that this approach allows teams to identify potential traps or biases in AI, and then openly address them. Perhaps the final takeaway from the First AI Day held by Health@WPP was a message of hope. When the right people are leveraging the right information and innovating without fear, there’s no telling what the market can do.

“Although it might take us time to get there, the starting point is the belief and the ambition that we want to make it happen,” said Alex Condoleon, chief medical affairs officer of Global Medical Omnichannel and Digital Customer Solutions at Pfizer.

“But the net value, I think, is immense. I hope we keep pushing that 10% disruptive frontier because I think there's huge societal benefit when we get there,” said Condoleon.

If you’d like to learn more about how enterprise AI can unlock huge value by solving your company’s toughest problems, contact Health@WPP.

The editorial staff had no role in this post's creation.