What challenges did Sanofi face before AI implementation?

Difficulty personalizing treatment for aggressive cancers (e.g., triple-negative breast cancer) with high mortality and resistance to standard therapies Traditional drug discovery takes 12-15 years, delaying life-saving treatments and resulting in patient deaths Lack of effective tools to analyze complex, multimodal patient data for biomarker and therapeutic target identification Need to identify suitable proteins/targets efficiently to expedite R&D pipelines Requirement to balance AI innovation with risk management and regulatory compliance

Source

MicrosoftExpandedEvidence: Low35/100

Sanofi personalizes cancer treatment and accelerates drug discovery using advanced AI

Use case typeTherapeutics researchUpdated Jun 13, 2026

Sanofi, a global pharmaceutical leader based in France, is leveraging advanced AI and Microsoft Azure technologies to transform two core areas: cancer treatment optimization and drug discovery acceleration. In partnership with AI-driven firms like Owkin, BioMap, and Exscientia, Sanofi employs large-scale federated learning, multimodal medical data analysis, and protein large language models (LLMs) to: 1) discover prognostic biomarkers and novel therapeutic targets in aggressive cancer subtypes; and 2) automate protein engineering and minor molecule candidate discovery for new drugs. Their Responsible AI framework guides transparent, ethical implementation. While most AI-enabled pipelines are recently launched or ongoing, the company has announced faster drug identification timelines, expansion into immunology, and progress of novel therapeutic agents to preclinical phases—all enabled by secure, scalable Azure cloud services. The implementation involves collaborations and integration between proprietary data, cloud AI services, and advanced analytics for precision medicine.

Organization: Sanofi
Industry: Pharma
Location: France
Published: October 2023

Planned next steps

The source says the organization aims to achieve: Enabled novel biomarker and target discovery.

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: Sanofi
Provider: Microsoft
Maturity: Unknown
Linked source: emerj.com

No explicit deployment-stage evidence found.

Customer identity supportedSource describes one deploymentMaturity evidence evaluated

Primary read

Use case focus

Showing 3 of 4

1Personalized oncology treatment optimization agent
2AI-driven drug discovery automation platform
3Federated learning for clinical biomarker identification

Difficulty personalizing treatment for aggressive cancers (e.g., triple-negative breast cancer) with high mortality and resistance to standard therapies
Traditional drug discovery takes 12-15 years, delaying life-saving treatments and resulting in patient deaths
Lack of effective tools to analyze complex, multimodal patient data for biomarker and therapeutic target identification
Need to identify suitable proteins/targets efficiently to expedite R&D pipelines
Requirement to balance AI innovation with risk management and regulatory compliance

Used federated learning and multimodal data analysis with AI models to detect new biomarkers and predict treatment response (cancer care)
Partnered with companies like Owkin, BioMap, Exscientia to access and develop AI-driven platforms and protein LLMs
Leveraged Microsoft Azure and compatible AI cloud services for secure, scalable compute and integration of proprietary data
Established Responsible AI principles to ensure ethical, robust, and transparent implementation
Automated workflows from biomarker discovery to molecule design, reducing time-to-market for drug candidates

Technologies

Azure AI Azure ML Azure OpenAI

Accelerated pipeline for up to 15 new minor molecule drug candidates in oncology and immunology

Architecture

Sanofi integrates advanced AI platforms (from Owkin, BioMap, and Exscientia) into its drug discovery and clinical trial optimization processes, leveraging federated learning on multimodal medical data via Microsoft Azure cloud. The process involves data access and curation, AI-driven patient and target characterization, and collaboration on secure, scalable Azure cloud infrastructure. Proprietary data and external AI models interconnect, automating workflows across biomarker detection, novel molecule candidate design, and patient stratification.

Sources & evidence1

Evidence: Low35/100Evidence strength