It's Friday, finally. Most of the week reaches us as separate events; the Long View is where we sit with one of them long enough to see its shape. Today that means a single company and a single wager, laid out in more public detail than most firms permit.
Start with two facts that look unrelated and aren't. In July, Sanofi walked away from amlitelimab, an atopic dermatitis candidate with positive Phase III maintenance data, on the judgment that it "would not represent a meaningful improvement to the standard of care." The decision carried a €952m impairment, and itepekimab, balinatunfib and a riliprubart study were cut alongside it. The same company, the same year, has been telling conference audiences it has taught itself to find drugs in half the time it used to. Those two things are halves of one strategy.
The Half That's Easy To Sell
Speaking at Bio-IT World, Michel Rider, Sanofi's global head of digital R&D, said the company has seen up to 50% efficiency gains in research cycle times after deploying AI tools and modernizing its platforms. A path that once took four years from target identification to clinical candidate now produces comparable output in roughly two. "We're able to generate the same output with reasonably quality molecules in half that time," she said.
The figure matches what Sanofi has been saying elsewhere. At BioAsia 2026 in Hyderabad, executive vice president and head of business operations Madeleine Roach said the company wants to be the first biopharma powered by AI at scale and to halve the journey from discovery to therapy. As Scrip reported, Roach pointed to specific proof: AI has cut mRNA design time by 50%, using in-house models including CodonBERT, mRNA-LM and RiboNN; the company's discovery engines produced seven novel drug targets in a single year; and Sanofi claims to be the first pharma company with AI-enabled trispecific antibodies for HIV and cancer in the clinic.