Several claims remain unverified. Bourla's account of the reversal rests on his own description, and no internal metrics have been published to show the new model is working. Independent analysis is skeptical. One outlet argues that the doubling of planned Phase 3 starts comes mostly from two 2025 acquisitions, not AI, and that AI's measurable contribution is on the cost side, trimming selling and administrative spending so R&D can grow. Pfizer's R&D burden has also grown with acquisitions including Metsera, 3SBio and Seagen, according to CFO Dave Denton, which complicates any attribution of pipeline gains to AI.
For attorneys, the story matters for three reasons. It is a rare public admission that a flagship corporate AI program underperformed, which may be cited in future disputes over AI investment claims, disclosures or governance. It also tests a broader industry pattern: Pfizer, Eli Lilly, Novartis, Bristol Myers Squibb and AstraZeneca are all expanding AI use, but drug discovery is not yet their clear success story. Watch whether Pfizer's shift to functional ownership, which bets that organizational design rather than technology was the bottleneck, produces disclosed results, and whether any future securities or product-liability filings characterize AI's role in pipeline performance.