Experts have theorized that artificial intelligence will fundamentally alter the pharmaceutical sector by increasing the efficiency of drug development and lowering costs. And while its potential is exciting, the current AI models have a long way to go before they disrupt medicine, according to pharmaceutical executive Taggart McGurrin.
That doesn’t mean AI won’t contribute to drug development. But it probably won’t create a revolution. At least not yet, McGurrin said.
McGurrin has helped lead innovative startups in the pharmaceutical space. But beyond his experience in the field, his training as a CFO/ COO and strategic executive has made him uniquely qualified to understand the industry from various important and often-overlooked angles.
When he thinks about the future of drug development, he doesn’t just consider the mechanisms that may improve the efficiency of the research and patient trials but the entire process that begins with the kernel of an idea and ends with new medicine hitting the market.
For fans of next-generation technology, the recent maturation of artificial intelligence has been a cause for celebration. Futurists predict that by harnessing the ability of AI software to comb through gigabytes of data meticulously, pharmaceutical companies could invent and test future medicine at a fraction of the cost of the current drug discovery model. All that’s missing is the right program.
But the pharmaceutical industry is complex. It requires rigorous, precise testing. Current AI models are prone to making mistakes or “hallucinating” false information, which could be truly dangerous in medicine.
Is it possible to make more innovative, more vigilant artificial intelligence programs that can be trusted to take the safety of human lives into account when they operate—or will AI be just the next step in technology, advancing the industry inch by inch?
How AI Can Help
While McGurrin urges caution about the world-flipping power of today’s AI platforms for drug development, he acknowledges its potential. Even the current artificial intelligence models can deliver big benefits to companies in the pharmaceutical industry, he said.
“It’s far from the first time AI has consumed headlines, but now you have LLMs [Large Language Models] and generative AI, both of which can fit squarely within the pharmaceutical industry so it can be more efficient and effective in drug development processes, thereby decreasing cost and time, eventually making it easier to bring down drug pricing,” McGurrin said.
AI tools have already shown the capacity to streamline and improve existing workflows, such as document management, in pharmaceutical companies.. What’s more exciting, however, is how artificial intelligence could play a role in reducing the inefficiencies within the drug development process, McGurrin said.
For example, the current systems for testing new drugs could be significantly improved by employing AI’s ability to manage large chunks of data. Before the federal government can approve new drugs, they must go through rigorous rounds of testing. Artificial intelligence could aid that process by sorting through data about potential patients and identifying those likely to experience complications or side effects from a new drug based on their health histories or current medications.
In addition, AI has the potential to reduce the time scientists spend theorizing how new compounds will interact with the human body. The same kind of technology that can distill vast troves of information into readable synopses could predict how new drugs will affect human biological systems.
It could also be a huge boon to research, helping scientists worldwide more quickly and easily find, translate, and understand the latest findings from cutting-edge papers across medical disciplines.
The Drawbacks of AI
But despite its potential, artificial intelligence may not significantly alter the pace of new drug development for a long time, McGurrin cautioned.
Even if generative AI programs can help medical researchers predict animal and human drug testing outcomes, identify more suitable patient populations for tests, and expand access to the latest research, the medical establishment will still move slowly. And that’s for a good reason.
“People think, okay, we’ll just get generative AI, and then prices will come down. But it’s more complicated than that,” McGurrin said. “The process is still costly, and there’s pressure to meet high standards—AI won’t just erase those expenses.”
It will be many years before humans take the word of artificial intelligence programs without doing much confirmation work. The stakes need to be lowered to leave to a yet-unproven technology.
In addition, the process that drugs need to go through to be labeled safe for consumers is necessarily long. Medical researchers must understand how a drug will work in the body over a lengthy period before they feel comfortable letting doctors prescribe it. It may be possible that side effects don’t show up until a few months of continuous use, meaning that drug trials (especially in phases II and III) need a long timeline to complete.
Modern pharmaceutical startups already work around the clock to bring drugs to market quickly, and it still takes a long time. “While AI could shave time off certain aspects, there simply aren’t enough areas to improve to speed up the overall process significantly,” McGurrin said.
“There’s an enormous demand to get drugs through the pipeline quickly, especially in biotech startups,” he said. “The cost of capital and the drive to deliver fast returns increases pricing pressure.”
Although he envisions a future where AI enhances the industry’s ability to address patient needs affordably, he stressed that we are “not there yet.” For McGurrin, AI will be transformative, but only if the pharmaceutical industry—and its investors and regulators—accept that it is just one part of a complex solution.
