Boehringer, Microsoft Explore AI Twins for Clinical Trials

Boehringer Ingelheim and Microsoft have announced a strategic collaboration to investigate how artificial intelligence (AI) and Digital Patient Twins could support more patient-centric and efficient clinical development. The initial work will focus on lung health, with the partners evaluating whether computational models can predict disease progression and treatment outcomes under different scenarios.

The collaboration remains exploratory. Its potential value will depend on whether the models can generate sufficiently reliable evidence to support clinical development without compromising patient safety or established regulatory standards.

Digital Patient Twins Anchor the Research

Digital Patient Twins are computational representations of individual patients built using clinical and other relevant data. Researchers can use these models to investigate disease trajectories and predict outcomes under different conditions.

The collaboration will examine disease-progression modelling, predictive outcome modelling and advanced clinical-trial simulation. A central objective is to assess whether these methods could support synthetic or hybrid control arms.

Synthetic control arms use appropriate existing patient data and analytical methods to construct a comparator group. Hybrid approaches combine existing data with a conventional control arm. If validated, these approaches could provide additional options for evaluating treatment effects in clinical trials.

However, their usefulness depends on the quality and relevance of the underlying data, the comparability of patient populations and the ability of the models to reproduce clinically meaningful outcomes. Differences in baseline characteristics, disease severity and other confounding factors could affect the reliability of the comparisons.

Lung Health Is the Initial Focus

Boehringer Ingelheim and Microsoft intend to assess the scientific and technical feasibility of their approach through modelling, prototyping and evidence generation. The work will begin with lung health before the partners evaluate its applicability across additional disease and therapeutic areas.

This initial focus provides an opportunity to test whether computational models can adequately represent disease progression and support meaningful clinical-trial simulations. The announcement does not establish that Digital Patient Twins are already validated for routine clinical-trial use.

The results will need to demonstrate sufficient reliability for their intended applications. Any proposed use in clinical development would also need to meet applicable regulatory expectations.

Patient Safety and Regulatory Standards Remain Central

Dr. Lykke Hinsch Gylvin, Chief Medical Officer at Boehringer Ingelheim, said the collaboration aims to make clinical research more responsive to patients’ unmet needs while maintaining scientific quality, patient safety and compliance with regulatory requirements.

These safeguards are particularly important when considering alternative control strategies. A smaller conventional control group could potentially allow more trial participants to receive an investigational treatment, but reducing the number of control participants is not automatically appropriate for every study.

The acceptability of synthetic or hybrid controls will depend on the disease being studied, the available evidence, the trial design and the ability to establish a credible comparison. The collaboration therefore represents an investigation into potential methods, rather than an established replacement for conventional control groups.

AI Ambitions Extend Beyond Clinical Development

The partnership also encompasses technology infrastructure, security and datacenter modernization intended to support AI applications. Boehringer Ingelheim has outlined ambitions to expand responsible AI adoption across Human Pharma and Animal Health, supported by employee skills development, governance and secure technology foundations.

Harsha Deshmukh, a member of Boehringer Ingelheim’s Board of Managing Directors responsible for IT and Global Business Services, highlighted the importance of enabling employees to use AI responsibly and extending relevant solutions across the business.

The broader collaboration places clinical-development research within a wider technology strategy. However, infrastructure modernization and employee AI adoption represent separate activities from the scientific validation of Digital Patient Twins.

What the Collaboration Needs to Demonstrate

The central question is whether AI-backed patient models can produce reliable predictions that improve clinical-trial design and evidence generation. Potential benefits include more informative simulations, more efficient use of existing clinical data and, where scientifically justified, reduced reliance on conventional control-group recruitment.

Those benefits remain prospective. The collaboration’s significance will depend on the quality of its validation work, the reproducibility of its predictions and whether the resulting methods meet the requirements for their intended clinical-development applications.

For now, the partnership signals growing interest in applying AI to clinical-trial methodology. Demonstrating that computational patient models can support credible treatment comparisons, rather than simply generate predictions, will be essential to establishing their practical value.

Reference

Boehringer Ingelheim and Microsoft partner to drive patient-centric and AI-backed clinical development, Boehringer Ingelheim, 22 October 2026