AI at heart of Roche’s future strategy

Roche presented its future plans to investors on Monday. Keystone-SDA

Swiss pharmaceutical giant Roche outlined its ambitious future plans, which include greater use of artificial intelligence to ​speed up research and development discovery, at ‌its pharma investor day on Monday.

Teresa Graham, head of the pharmaceuticals division, said the Basel-based company had many promising potential blockbusters in its pipeline. While Roche was hit by research setbacks and revenue losses due to expiring patents in 2022, the group has got itself back on track thanks to an ambitious programme, she told attendees at Roche’s “Pharma Day” on Monday.

“Our track record since last year’s “Pharma Day” underscores our progress. Since then, we have achieved our targets in ten regulatory-relevant trials and advanced six programmes to the regulatory phase, with a further three entering Phase II,” she said.

Roche has up to 20 new active substances that could reach the market by 2030. Some, such as the breast cancer drug giredestrant, have the potential to generate peak sales of more than CHF30 billion ($36 billion).

What’s the state of the pipeline?

Roche introduced “The Bar” at the end of 2023 as a new evaluation framework for its research pipeline. Since then, drug candidates have had to meet certain criteria in order to proceed to further development.

Roche’s chief medical officer, Levi Garraway, stressed: “‘The Bar’ has become the guiding principle for our research and development.”

The selection process for potential candidates has become significantly more rigorous. “We have reduced the number of projects whilst significantly increasing the overall value of our remaining portfolio,” he said. The result is that the success rate for projects reaching the crucial Phase III of clinical development stood at 65% in 2025 and has so far been more than 80% in 2026. At the same time, development times have been shortened and costs reduced.

Roche and AI

Artificial intelligence is playing an increasingly important role for the pharma group. Roche does not simply want to use AI for individual analyses, but to integrate it as early as possible and throughout the entire drug discovery process, from understanding a disease, through the search for new drug targets, to the selection and optimisation of drug candidates.

Aviv Regev, Head of Genentech’s Research and Early Development (gRED) department, presented “Lab-in-the-Loop”. AI models analyse large volumes of biological and chemical data, make predictions based on this, and these are then verified in the laboratory. The new experimental results are in turn fed back into the models.

“It is important that our confidence in the reliability of these models grows over time, but AI cannot replace clinical trials in humans,” he said.

How we produce news in English

Translated from German, reviewed by an English Department journalist.