The British startup Exscientia claims it has developed the first medication created using artificial intelligence that will be clinically tested on humans. The medication, which is meant to treat obsessive-compulsive disorder, took less than a year from conception to trial-ready capsule. Human trials are set to begin in March, but would you take a drug designed using artificially intelligent software?
The appeal of AI-designed drugs is relatively straightforward. There are lots and lots of possible molecules that might be useful in medications, far too many for all the medical researchers in the world to manually test. But by using different types of AI, a computer system can come up with and mine through different molecules, comparing them against different parameters and learning the most promising compounds faster than a human could.
Artificial intelligence is certainly powerful, but some are skeptical that the technology is reliable or trustworthy, and question what role it should play in areas like our health care. And in drug research, some have voiced concerns that the technology might be overhyped, and that AI’s findings may not be as groundbreaking as we’d like to believe.
Exscientia’s CEO, Andrew Hopkins, has argued that AI means synthesizing fewer compounds for testing and running fewer experiments in the search for a new medication. “Active Learning [a subcategory of a type of artificial intelligence called machine learning] algorithms automatically prioritise the most informative compounds for experimental synthesis and testing and enable the system to learn faster than humans alone,” Hopkins told Recode in an email.
Of course, AI could be used for more than just trying to come up with new compounds. The technology could also mine through scientific research and patient data and help inspire the repurposing of old drugs, among a wide variety of other applications. And the uses can be even broader than just medication: Researchers have used AI to track the spread of the Wuhan coronavirus, and the technology is also being deployed to tackle the American opioid crisis.
Hopkins went on to explain that his company’s platform is the first to have produced a drug with artificial intelligence that will be tested in a clinical trial. He says the compound that was ultimately produced — called DSP-1181 — is expected to last longer and have a stronger efficacy than other existing medications for OCD. Japanese pharmaceutical company Sumitomo Dainippon Pharma, which owns the rights to the drug, will oversee its clinical development. The first phase of human trials, which will test the medication’s safety and the body’s response to it, will also take place in Japan.
As remarkable as the new drug’s development seems, there’s still room for some healthy skepticism. Artificial intelligence can help us find new molecules, but there’s a chance that the molecules found by AI will ultimately resemble molecules we’ve already studied. That’s the caution of chemist Derek Lowe, a researcher at Novartis who works on drug discovery. On his pharmaceutical industry blog, Lowe explains how simply finding a potential compound doesn’t guarantee that scientists actually understand the biochemical nature of the illness they’re trying to treat — or that the drug will even work.
“The problem is that preclinical drug optimization is not the problem,” he wrote of Exscientia’s announcement. “This project, at best, seems to me to have saved a few months off the process of sending their compound into the same black-box shredder as every such drug project goes into when it hits human trials.”
Meanwhile, the development of AI-assisted pharmaceuticals raises questions about how comfortable people should be with these new research methods. In the long run, how will AI-designed drugs differ from those developed by humans alone? Who should make the rules for the use of AI in drug research?
Like all applications of artificial intelligence, health authorities are trying to figure out how best to study and regulate these tools. While the Food and Drug Administration wouldn’t comment on this particular new drug, FDA spokesperson Jeremy Kahn told Recode that the agency is committed to maintaining public health standards — while also protecting innovation — and that its drug evaluation and research center is assessing the regulatory considerations that AI tools might raise.
“The full role of AI in drug development is still being elucidated, and stakeholders understand AI in different ways considering the spectrum of tools and techniques covered under this umbrella term,” Kahn told Recode in an email. “Importantly, the evidentiary standards needed to support drug approvals remain the same regardless of the technological advances involved.”
Meanwhile, a spokesperson for Exscientia said the drug had to meet the same requirements as any other drug undergoing a phase I trial in Japan.
It’s essential to remember that Exscientia and other pharmaceutical companies stand to make a lot of cash if AI-based drug development actually works. The number of investments major biotech investments are dropping on the technology suggests as much. Large pharmaceutical companies are increasingly throwing money behind artificial intelligence, and Exscientia, whose investors include the German drug company Evotec and Bristol-Myers Squibb, is working with several drug giants on new drugs, including Bayer and GlaxoSmithKline.
So you might not have the opportunity to try a drug designed by artificial intelligence in the next few months — unless you’re one of the select humans that will try this new drug in Japan. Still, this recent development brings us closer than ever to a future where AI designs new drugs. Will they actually aid our ailments any better than good old-fashioned human-made drugs? We’ll have to wait and see.