As the marriage between pharma and digital health heats up, intellectual property and patent disputes over artificial intelligence and other nascent technologies could lead to a flurry of lawsuits, experts say.
Pharma is increasingly partnering with digital health companies for drug discovery, patient engagement, clinical trial automation, manufacturing and supply chain operations, and more. But experts say the lack of clear regulatory guidance over data ownership could lead to disputes in court.
“I expect litigation to arise in situations where a digital health company felt that a large institutional player used them and took their proprietary information, got enough out of them and then hung them out to dry,” said Shabbi Khan, intellectual property lawyer with Foley & Lardner. “Younger startups have less leverage against these bigger players. The protection lies in the agreement you've established between [the two parties].”
Anne Li, an intellectual property lawyer at Crowell, is even more frank. She said patients, digital health companies and pharma could be involved in suits over privacy and data ownership as these technologies become more commonplace despite gaps in regulatory guidance. Li expects these issues to be debated on three fronts: within regulatory bodies, in courts and legislatively.
“When it comes to regulatory guidance, I would say the most disturbing thing I’ve heard is there isn’t one specific area that needs to be addressed,” Li said. “Everyone I’ve talked to says are a lot of issues that need to be addressed. If it were one thing, the regulatory agencies and legislature could probably work on it, but that’s not the case.”
Data is not patentable
Since 2021, there have been 321 filings for AI-related patents in the pharma space, according to research firm Pharmaceutical Technology, nearly as much as the previous six years combined. But while digital health companies can patent their initial algorithms, the U.S. Patent and Trademark Office has ruled AI cannot be an inventor of new patents.
This means once a company’s algorithm starts generating novel insights on its own, it cannot be patented and protected. The U.S. Eastern District of Virginia ruled in favor of this finding in September.
“Data is not patentable,” Li said. This doesn’t mean digital health companies are out of luck, she said, they just need to protect themselves. “When you're entering into joint agreements or doing clinical trials, it's very important to have all the correct structures in place to protect patient confidentiality and your dataset.”
Li said agreements between digital health companies and pharma can age out very quickly as technology rapidly evolves. She said digital health companies should be filing amendments over data ownership and trade secret protection when necessary. Anything older than three or four years old probably doesn’t have the right provisions, she said.
Khan said digital health companies should add conditions that compensate them for any insights generated by the algorithm after it was launched. He also said that they should always assume that the partnering company is likely watching out for their own interests and filing for protection on future ideas.
“If you’re partnering with someone else, you need to continue to protect your IP that you are generating outside of that relationship,” Khan said. “If you ever become dispensable to them, your IP that you own separately can help either bring them back to the table or restrict them from partnering with somebody from a with a competitor.”
Nick Mitrokostas, an intellectual property lawyer at law firm Allen & Overy, said that the industry is figuring out ownership rights when more than one company has access to data, particularly coming from an algorithm that’s learned new insights. What makes things even more confusing is that different countries have come to different conclusions on AI’s role, he said. For instance, South Africa became the first country to accept that AI could be an inventor last year.
“I think the regulatory bodies, patent, drug approval and medical device agencies in various countries are trying to figure out how this disruptive force in the market has changed their traditional approach to the approval of a new drug, medical device or issuance of a new patent,” Mitrokostas said.
Khan said that digital health companies must ensure their tech stack is flexible enough to set satisfied various requirements or different jurisdictions. “If don't design your platform to be scalable and to be accommodating across jurisdictions, that’s bad design,” he said.