As drug candidates discovered via AI move into later-stage clinical trials, the technology seems to be doing as promised: speeding drug development.
The early trial results are in, and AI appears to be doing what was promised: speeding clinical trials and improving biotech success rates. This is a boon for biotechs that got in early and validation of the sky-high valuations they have enjoyed, according to PitchBook.
“Fundamental advancements in drug discovery are nothing short of revolutionary,” PitchBook wrote in a year-end report in December that specifically emphasized AI. The firm went on to predict that AI will help “nearly double” investigational new drug applications success rates while also cutting development timelines and costs.
If drug development is faster, venture capital firms can put their cash back into new companies sooner; it’s a win-win for the entire biotech ecosystem, PitchBook wrote. “Faster validation cycles could meaningfully improve capital recycling in a segment historically constrained by long development timelines and limited early liquidity.”
AI can particularly help early-stage companies, which have struggled to get VC dollars over the past year, because AI-derived trial improvements are more prominent at this stage, PitchBook wrote.
PitchBook’s early analysis showed that AI-native biotechs, which are companies that use AI as a foundational technology, have so far achieved about an 80% to 90% Phase I success rate, compared to the industry average of 40%–65%. The success rate dropped to just 40% in Phase II, but that still tops the current industry average of 29%. PitchBook noted that AI-focused biotechs have completed just 10 clinical trials so far, however.
“Although the dataset is nascent, the higher Phase I success rates may reflect improved target selection,” PitchBook wrote.
As one example, Insilico’s investigation of 22 drug candidates took just 12–18 months to go from project initiation to nomination of preclinical program, shortened from the typical length of 2.5 to four years, according to ICON Global’s 2025 global biotech sector survey report.
While PitchBook did not provide specific numbers, the report noted that the AI revolution is expected to boost return on investment potential as compared to the 2012 to 2014 vintage, which provided life sciences venture capital investors with a 19.5% internal rate of return.
The world will have to wait as these assets move into later-stage development to see if the overall success rate improves. But PitchBook thinks it will: “Should these early results persist as we expect, this preliminary data suggests the probability of success may increase from approximately 8% to 18%.”
Party Like A Pharma
AI is also helping biotechs act more like Big Pharmas. PitchBook said these companies “can produce more shots on goal without escalating costs—a structural de-risking.”
Julia Tarasenko, chief commercial and strategy officer at lab services company LabConnect, said that AI is helping these small companies run better trials.
“This means new treatments can reach patients much quicker, and smaller companies can now compete with big pharma,” Tarasenko told BioSpace in an email. “In the long run, this could dramatically lower the cost of drug development and make trials more accessible for everyone involved.”
On the flip side, integrating AI can be tougher for bigger companies, according to Graham Mills, a principal at London-based life sciences venture fund Abingworth who was quoted in ICON’s survey.
“Even if they perceive that it will have an impact, scaling usage within a large bureaucratic organization can be difficult so we may see different uptake from that,” Mills said.
Sanofi is one bigger company that has gone deep into AI, with partner Exscientia becoming now eligible for €600 million ($703 million) in milestones in 2024 alone, an early example of the potential for swift growth out there for AI-focused startups.
Indeed, the technology is going well beyond boosting clinical trial efficiency. Other applications include general operational efficiencies and drug discovery. ICON said that AI can help as biotechs face a more constrained labor market and the hunt for talent gets even fiercer.
AI could even make M&A more efficient, according to Greg Graves, senior partner in McKinsey’s life sciences practice. Companies with well-integrated AI infrastructure could more quickly integrate new companies—which means bigger, more complex deals will be less daunting to take on.
But companies have to have the talent to run and understand the tech, too. “The willingness to adopt AI tools depends on a company’s risk appetite, and having the budget and skills to create AI solutions in house or to source them externally,” ICON said.
