By Allison Proffitt 

November 12, 2024 | Last week at SCOPE Europe, industry leaders took on the impact of AI regulation—particularly the European Union's AI Act—on the clinical research and life sciences sectors. 

Moderated by BioSpace’s Lori Ellis, the panelists included Firas Abdessalem of Sanofi, Christopher Hart of Foley Hoag, Ricardo Gaminha Pacheco of InSilico Medicine, and Artemy Shumskiy of Latvian VC company LongeVC. Together they discussed how the EU’s AI Act is likely to impact innovation, funding, and partnerships in the life sciences.  

One of the primary concerns voiced by panelists was the potential stifling effect of stringent AI regulations on innovation and the formation of new companies. "I think generally regulation cycles develop in the market, and especially regulation that is region-specific. I'm guessing that across the whole market, you will see a decrease in the number of companies or new companies formed," said Shumskiy. 

This sentiment underscores a broader apprehension that heavy regulatory burdens may deter startups and smaller enterprises from entering the AI-driven healthcare space. The life sciences industry has experienced something similar with GDPR, observed Abdessalem. While large corporations might weather these changes, smaller businesses could find compliance costs prohibitive, potentially leading to a consolidation of market power among established players, he said.  

Regional Disparities and Regulatory Outlook 

The EU’s AI Act highlights existing regional differences, the panelists agreed. Europe and Canada are perceived as taking a more cautious approach to AI tools than the United States and China, and the panel believes that these regulatory approaches will change the types of tools companies develop.  

"Founders will just not bother developing AI solutions or will go to regulatory states that are more lax in terms of those regulations," Shumskiy predicted, suggesting that companies might pivot toward regions with more favorable regulatory environments. 

This shift could have significant implications for investment, particularly from non-European investors who may be hesitant to support ventures grappling with complex regulatory landscapes. "Investors, especially outside of Europe, will not be as inclined to support companies in the space of developing AI solutions," he predicted, highlighting a potential slowdown in funding for innovative projects within the EU. 

“I feel like we’ll see a divide between countries that are trying to compete on AI, such as the [United] States and China, and countries that are trying to regulate it away so that people are less impacted, such as EU and Canada,” he said.  

Chris Hart, from US law firm Foley Hoag, argued that in the US, AI-specific regulation is unnecessary. “In the US—not unlike any other market we’re talking about—anything having to do with drug development, devices, patient care, clinical trials—it is already a heavily regulated space. A lot of the things we’re talking about here strike me as already well-covered with existing regulations.” It’s an argument he says is often echoed by US lawmakers. “We have a web of different kinds of laws that are going to allow law enforcement or regulators to do what they need to do.”   

InSilico Medicine’s Pacheco agreed. “When we have something, especially a compound that’s going to be developed for a particular disease that somehow AI has touched, the level of scrutiny from the laboratories evaluating it is even more than it generally is.”  

Impact on Drug Discovery and Compliance 

There was optimism about specific niches within the healthcare sector. For example, Pacheco expects that the regulation may not be as impactful across the board as some fear. “There’s an exemption, I believe, with the AI Act for tools for software,” he said. “From our perspective,” he added, “we do not expect a major disruption because our tools are mainly focused on early drug discovery, where there's low risk.” This freedom for low-risk areas could allow for continued innovation without companies bearing the full brunt of regulatory constraints. 

Pacheco did note that while his company may be somewhat insulated, for the pharma and biotech companies with which he partners, complexity and stringency of compliance will likely increase, but he reported not seeing any detrimental impact of the Act on his own dealmaking or assets-licensing.  

Steps Forward 

While the panelists agreed that the largest companies—Google and OpenAI were mentioned as examples—probably won’t have trouble penetrating the EU market or managing the demands of the AI Act, small to medium enterprises, biotechs, and smaller pharma should be prepared.  

Hart recommended organizations be agile as different regulatory landscapes change. “What we should do is watch what the enforcement authorities do. Watch how they interpret it. Watch what guidance says. Ultimately, that should give you direction.”  

Abdessalem added that organizations should also pay attention to what more “lax companies”, are doing.  

Pacheco suggested that smaller companies might prefer to start their businesses in more regulatory-friendly locations before expanding to the EU. This is a trusted approach, Shumskiy said. “In the healthcare space, especially in drug discovery and drug development, diagnostics, and pretty much any deep-techy healthcare space, everything happens in the [United] States anyway,” he said. While, of course, clinical trials are a global endeavor, “the first market that pharma considers… is typically the States,” he noted.   

But Abdessalem warns these small organizations outside of the EU not to ignore the current conversation—especially ones that hope to be acquired or expand into more heavily regulated areas. He recommended establishing a responsible AI framework as soon as possible. “I think it’s harder to reverse engineer a validated system,” he said.