How a Basel Biotech SME Uses AI to Reduce Animal Testing

Swiss biotech SMEs face mounting pressure to deliver safer, more effective medicines while meeting ethical and regulatory expectations. In Basel, one SME has adopted AI-driven simulation to reduce animal testing in preclinical drug development—achieving faster results, lower costs, and improved compliance with evolving standards.

The Challenge: Ethical and Efficient Drug Development

Drug discovery and early testing are among the most resource-intensive activities for biotech companies. Traditionally, before advancing a candidate drug to human trials, SMEs must run extensive animal experiments to demonstrate safety and efficacy. This process is costly, slow, and increasingly scrutinised by regulators and the public alike.

For a Basel-based SME specialising in RNA therapies, the high volume of required animal tests was a bottleneck. The company needed to:

• Reduce time and cost spent on preclinical animal studies
• Increase confidence in advancing only promising drug candidates
• Strengthen compliance with Swiss and EU ethical standards for animal welfare
• Maintain competitive pace against larger, better-resourced rivals

The Solution: AI Foundation Models for Preclinical Simulation

Inspired by recent advances like those from Atlas Bio, the SME partnered with a Swiss AI provider to deploy an AI foundation model trained on millions of RNA readouts. This model was designed to simulate patient-level efficacy and predict potential safety issues for new RNA-based drug candidates—before any animal testing was performed.

Key Features of the AI Solution:

• Data-Driven Predictions: The AI model analyses large-scale RNA datasets, historical lab outcomes, and chemical properties to predict how a new candidate might perform.
• Virtual Patient Simulation: Generates simulated responses across diverse virtual patient populations, identifying likely efficacy and risk profiles.
• Early No-Go Decisions: Candidates unlikely to meet safety or efficacy thresholds can be deprioritised, avoiding unnecessary animal studies.

Implementation Steps: Rolling Out AI for Preclinical Insights

1. Data Preparation:

   • Internal lab results, anonymised clinical data, and chemical structures were securely gathered.
   • Data quality was assessed and harmonised with support from the AI vendor, ensuring compatibility with the foundation model.

2. Model Customisation:

   • The off-the-shelf AI model was fine-tuned using both public and proprietary RNA and drug outcome data.
   • SME researchers worked alongside AI specialists to calibrate predictions for the specific drug modality.

3. Simulation and Validation:

   • For each new drug candidate, the AI simulated patient-level outcomes for efficacy and common safety endpoints.
   • Results were benchmarked against historical animal studies to validate predictive accuracy.

4. Decision Integration:

   • The SME established new review checkpoints: only candidates passing AI-based simulation were advanced to animal studies.
   • Documentation of AI predictions was incorporated into regulatory submissions, demonstrating due diligence and ethical commitment.

5. Continuous Learning:

   • New animal and human trial data was fed back into the AI model, continually improving its predictive power.

Tangible Benefits for the SME

• Reduced Animal Testing: The company estimates a 20–30% decrease in animal experiments for early-stage candidates.
• Faster Pipeline Decisions: Simulations provide go/no-go insights weeks ahead of traditional workflows, accelerating project timelines.
• Cost Savings: Fewer animal studies mean significant savings in laboratory resources and staffing.
• Regulatory Alignment: Transparent use of AI supports compliance with Swiss and EU ethical standards, and strengthens trust with regulators.
• Competitive Advantage: The SME has been able to advance high-confidence candidates more quickly, improving its value proposition to partners and investors.

Lessons for Other Swiss SMEs

This practical case shows that AI-driven simulation is no longer limited to global pharma giants. Swiss biotech SMEs can tap into advanced foundation models—either through partnerships or by leveraging recent Swiss AI innovation—to reduce ethical burdens and enhance R&D efficiency. The key is to start with well-organised data, collaborate closely with AI specialists, and integrate predictive insights into formal decision-making processes.