While the industry often approaches artificial intelligence (AI) to decrease drug cost and development timelines, the value of AI within biopharma is its capability to deliver first-in-class or best-in-class therapies, said Atomwise CEO Abe Heifets at the 2024 BIO CEO & Investor Conference in New York.

A treatment developed with or without AI is valuable to the industry if it can improve upon the existing therapies, Heifets explained. In this sense, AI can act as a “toolkit” to expand what the industry can already do, added Liz Schwarzbach, chief business officer of BigHat Biosciences. Both were speaking at a panel on the value of AI at the conference, which closed on 27 February.

“AI is an enabler; it is an extension of capabilities,” said Zachary Carpenter, CEO of VantAI, a company that recently signed a drug discovery deal with BMS worth $674m.

While AI is being used to get drugs to the clinic quickly, Heifets said the important question is if those investigational drugs meaningfully differentiate versus the precedent once they are trialled. “Patients don’t benefit when drugs get to the clinic; patients benefit when they get through the clinic,” he added. Atomwise is a San Francisco, California-based startup using its proprietary AI technology to aid drug discovery.

The panel also saw experts touch upon the concept of technology first biotechs or “techbio” companies. The purpose of AI is to deliver robust data that propels drug discovery, said Petrina Kamya, VP and global head of AI platforms at Insilico Medicine, another AI-focused company. Having a techbio structure ensures robust data is collected by leveraging deep learning algorithms to discover novel targets. This approach utilises generative AI to design small molecules that are optimised for certain properties, and also allows a company to prioritise indications, and

There are fundamental differences between tech bios and biotechs. Techbios are a “loop,” where every byte of data is viewed as value and the data is tested, validated, and trained, said Carpenter. This approach is a shift from biotechs where a lot of data is wasted on capture, he added.

The strength of techbios is that they can continue creating value for patients, and for longer periods of time because the knowledge remains intact in the company.

All things being equal, it’s understandable to want to get a drug out faster and cheaper for a patient population, Heifets told Pharmaceutical Technology after the panel. The implication is that as an industry, “we can’t conceive things being better; and we need to think about better,” said Heifets. The hypothesis is that tech bios can generate the value the industry needs.