Pharma’s AI prospects get nudged into the future with EU’s AI act
Caption: Experts at the LSX World Congress 2025 discussed the potential challenges and benefits associated with the EU AI Act. Credit: Leestat via Getty Images. ·Pharmaceutical Technology·Leestat via Getty Images.
Akosua Mireku
6 min read
The EU Artificial Intelligence (AI) Act may protect citizens, but some experts in the life sciences industry are worried about the potential challenges it may raise for business.
As AI continues to permeate the drug development pathway, major players like Eli Lilly, Sanofi, and BioNTech have incorporated the technology into their development pathways through major deals. Meanwhile, NVIDIA has also forged several partnerships, including some with smaller companies like Gain Therapeutics, to propel their AI platforms, as well as therapies, to major market success over the next few years. Gain’s work with the technology giant under the NVIDIA Connect program aims to propel its Magellan platform for drug discovery, says the company’s CEO Gene Mack.
Alongside its potential in expediting drug discovery, AI has been explored in the life sciences sector for patient recruitment, clinical trial design, and more, but some still critique the lack of regulation surrounding the technology. At the LSX World Congress, held 29-30, experts shared their thoughts on the potential outcomes of the EU AI Act and future directions for AI regulation.
The EU AI Act (Regulation (EU) 2024/1689 ("EU AI Act"), first published in the EU Official Journal on 12 July, 2024, is said to lay out a comprehensive framework as the first regulation on AI worldwide. The Act classifies use of AI by risk levels: unacceptable, high, limited, and minimal risk. Unacceptable risk includes AI systems that manipulate people to engage in unwanted behaviours and real-time and remote biometric identification systems, among other things, while minimal risk covers things like AI-enabled video games and spam filters.
The AI Act was first implemented in August 2024 and will fully come into force in August 2026, excluding certain exceptions. For example, restrictions and rules surrounding AI literacy were enacted as of 2 February, 2025, and governance rules and the obligations for general-purpose AI models became applicable on 2 August, as per the European Commission’s website.
As per a GlobalData report on AI, an organisation that uses AI systems would be subject to fewer obligations than a provider or an organisation that develops, markets, or supplies AI systems or models under the EU AI Act. Companies developing high-risk AI systems have until 2 August, 2027 to follow the new guidelines. Furthermore, businesses that fail to comply risk penalty fees of up to €35 million or 7% of their global revenue.
The regulation should “help us protect our fundamental rights,” said Leon Doorn, the chair of medical devices and AI at NEN, the Dutch Standardisation Institute, at the LSX conference. However, concerns remain surrounding how the act will align with existing regulations. “From a legislational level, the EU AI Act and the Medical Device Regulation (MDR) connect well, but on the regulatory side, there is complexity,” Doorn added.
Crossover with the Medical Device Regulation
The EC first passed the Medical Device Regulation (MDR) in 2017, with a staggered timeframe for full enforcement of the law set for 2027. The regulation updated medical device policies set out in the Medical Devices Directive (MDD) of 1992. Currently, under the MDR, commercialised medical devices must qualify for CE marking, ensuring that the legal manufacturer has assessed the device and confirmed that it aligns with general safety and performance requirements under the MDR. “One of the best things the AI Act is doing is aligning with the MDR,” said Fernando Benito, the chief technology officer at Custom Surgical, a Munich, Germany-based medical device company.
Pharmaceutical companies that are compliant with the AI Act may benefit from the same positive image that the CE mark has given businesses, said Doorn. Regulation may “stifle” AI development in life sciences, but the additional guidelines may instill increased trust for consumers, added Sridevi Nagarajan, former global head of digital regulatory strategy at AstraZeneca, during the conference.
However, moving from enacting the MDR to the EU AI Act raises questions about how regulatory bodies can make sure they do not “block progress,” said Bonito. Doorn said it was “frustrating” that the risk-based approach for the AI Act is not present in the MDR’s list of harmonised standards, so the combination of the two regulations may cause companies to shoulder a bigger burden.
Future possibilities
The new regulation’s risk categories may present some benefits for the industry, says Nagarajan. “Median [limited] risk AI may be able to leverage privileges,” added Nagarajan. The regulations around the limited and minimal risk level categories may instill more trustworthiness in pharmaceutical companies’ use of the technology for investors and users, she explained.
“People who are about to step into AI have an advantage [when following the EU AI Act], but things will look different for those already working with AI,” said Bonito. The new AI regulation may also be a greater challenge for smaller companies and startups, explained Bonito. “It [the EU AI Act] is only applied in Europe. Also, the MDR and General Data Protection Regulation (GDPR) already exist in the EU, so if you are now adding this on top, it may make European startups reach the market later,” said Bonito.
“American companies may reach the market sooner with, say, the fourth version of their AI technology, while European companies are still on the first version,” said Bonito. “The US companies can then remodel their technology for the EU regulation,” he added.
As per a GlobalData report on AI in Healthcare, the differences in AI regulation between countries may restrict the ability to train AI systems at a global scale.
On the other hand, some experts suspect that an earlier EU launch may be preferred despite the additional hurdles raised by the AI Act. “When the GDPR was published, it was only European, and now it’s worldwide,” said Doorn. Thus, if the EU AI Act follows a similar path, companies that comply sooner will see the benefits, he explained. “There is a great opportunity here for Europe to shine and not wait for America to move ahead,” said Nagarajan.
Problems with unregulated devices will begin occurring more and more frequently, so “it is time for countries and companies to embrace more AI regulation now,” added Nagarajan. “The EU AI Act is setting a precedent for the rest of the world…. Seeing that the EU is putting a big focus on the citizens, even if it puts a bit of damage on the industry, it’s a good thing,” said Bonito.
"Pharma’s AI prospects get nudged into the future with EU’s AI act " was originally created and published by Pharmaceutical Technology, a GlobalData owned brand.
The information on this site has been included in good faith for general informational purposes only. It is not intended to amount to advice on which you should rely, and we give no representation, warranty or guarantee, whether express or implied as to its accuracy or completeness. You must obtain professional or specialist advice before taking, or refraining from, any action on the basis of the content on our site.
