AstraZeneca CEO Pascal Soriot said the company is deploying artificial intelligence to de-risk late-stage clinical trials and accelerate regulatory filings, even as AI-designed drug candidates have yet to reach clinical testing. The disclosure came during the company’s second-quarter earnings call Monday.
Soriot acknowledged that AI has not yet produced any clinical-stage candidates from AstraZeneca’s internal discovery pipeline. However, he emphasized that the technology is already delivering value in later development phases by analyzing trial data to predict safety signals and identify patient subgroups most likely to respond to experimental therapies.
“There is huge potential for AI to impact the entire drug development continuum,” Soriot told analysts. He pointed to AI-driven biomarker analysis and real-world evidence integration as areas where the company sees near-term gains in trial efficiency and submission speed.
The approach mirrors a broader industry trend. Large pharmaceutical companies including Roche, Bristol Myers Squibb and Novartis have all announced AI partnerships this year aimed at improving late-stage trial success rates, which historically hover around 50% for Phase 3 studies. AstraZeneca’s strategy focuses on applying AI to existing trial infrastructure rather than waiting for AI-designed molecules to mature.
The CEO’s remarks come as AstraZeneca navigates pipeline challenges. The company recently reported disappointing data in cancer and rare disease programs, intensifying pressure to find efficiency gains through computational approaches.
