Insilico CEO Alex Zhavoronkov explains how machines are accelerating drug discovery and shaping longevity science.
What if the key to living healthier for longer isn’t just medicine but intelligence? Not human intelligence, but artificial intelligence (AI), quietly reshaping how drugs are discovered, tested and scaled.
In the latest episode of Longevity.Technology UNLOCKED, Phil Newman and Dr Nina Patrick sit down with Dr Alex Zhavoronkov, CEO of Insilico Medicine, to explore how AI is accelerating longevity biotech and why the next generation of healthspan therapies might arrive faster than we ever imagined.
Zhavoronkov’s philosophy is deceptively simple: knowledge before intervention. “Bet on diagnostic techniques because nowadays those techniques have advanced reasonably well and we can identify problems before they become deadly,” he says.
In practice, this means comprehensive screenings – think full-body MRIs, non-invasive blood tests and personalized medication profiling – to create a complete “data room” for each patient. It’s about preparedness. When disease arises, the more information doctors have, the better the outcomes. And increasingly, those “doctors” might include AI-powered systems like ChatGPT, Grok or Gemini.
From diagnostics, the conversation turns to AI-driven drug discovery, the beating heart of Insilico’s work. Zhavoronkov describes the company’s platform as a series of interlocking systems that address critical stages of the drug pipeline.
Biology 42 identifies promising targets and predicts which ones can be both effective and commercially viable. Chemistry 42 generates small molecules with precision. Medicine for You predicts clinical trial outcomes, and Science for You applies AI to sustainable chemistry, carbon capture and even agriculture.
“For Chemistry 42, that’s a system that allows you to very rapidly design small molecules with the desired properties. So instead of looking for a needle in the haystack, you generate perfect needles,” he says.
The results are transformative. Traditional drug development from target discovery to preclinical candidate can take four to five years (or longer). Insilico compresses that timeline to roughly 14 months.
Zhavoronkov likens it to designing a high-performance car: “It’s kind of like developing a car, a Ferrari or a Tesla, where you are trying to go from zero to 100 very, very quickly.” Yet he is careful to note that once a drug reaches the preclinical candidate stage, regulatory rules slow the pace. “After that, it’s a super-regulated process. It’s kind of like driving on a highway with a lot of traffic signs. You cannot really jump over.”
AI, for Zhavoronkov, is not a replacement for rigorous testing; it is a tool to sharpen it. Insilico runs thousands of predictive benchmarks per program, tracking every molecule through multiple iterations and across several species, including monkeys, to gather long-term safety data.
“Basically, this year, 2026, I call it the year of benchmarks. So we created up to a thousand benchmarks per program,” he says. The ultimate goal: reduce animal and human experimentation in the future while building an enormous data library that informs safer, more effective therapies.
Selecting the right targets is both art and science. Insilico looks for dual-purpose targets – molecules that could treat specific diseases and simultaneously modulate aging pathways.
“GLP-1 is a Kardashian of all targets, right? We’re trying to find the next mTOR, the next NLRP-3,” Zhavoronkov says, using cultural shorthand to highlight a scientific ambition. These targets are exciting and practical. They must be safe, scalable and applicable across multiple therapeutic areas, from chronic inflammation to muscle wasting.
Zhavoronkov’s approach is refreshingly grounded in real-world constraints. He stresses the importance of human trials for evaluating long-term safety: “For most of my drugs, that’s exactly the pathway we want to take: approve for a disease, expand indications beyond the disease and then see if it may work in aging.” AI accelerates discovery, but it does not replace the rigor, ethics and unpredictability of biology.
There’s also a commercial dimension to consider. Using GLP-1 therapeutics as an example, Zhavoronkov notes the balance between blockbuster success and broad accessibility. As patents expire and competitors enter the market, companies must innovate further while lowering costs. The result? A feedback loop where investment, competition and AI-driven discovery push the field forward, expanding the possibilities of longevity medicine.
In this episode, Insilico emerges not just as a drug developer but as a systems thinker, blending biology, chemistry and AI into an integrated pipeline. Zhavoronkov frames longevity as a practical, infrastructural challenge, noting that it isn’t about chasing immortality but maximizing healthy years, creating therapies that are both effective and safe over a lifetime.
“The molecule is done. You can personalize it, you can select better patient populations, but the learning, the iteration, all happens before preclinical candidate nomination,” he explains.
As AI reshapes biotech, the future of longevity looks less like a distant dream and more like an accelerating, data-driven reality. With rigorous science, visionary engineering and an eye on practical outcomes, the next era of healthspan innovation is arriving faster than most of us expected.
Catch new episodes of Longevity.Technology UNLOCKED every Monday, where innovators, scientists and visionaries reveal what’s next in aging and longevity. Listen on Apple Podcasts, Spotify and YouTube.
