Regulatory principle:

AI should be designed and used only within a clearly defined context of use, appropriate to the regulatory or scientific question being addressed.

How we are aligned:

AI in PreQlinic is applied to bounded, pre-specified functions, including:

• Structuring preclinical planning logic
• Mapping regulatory expectations to development activities
• Identifying risks, dependencies, and sequencing constraints
• Supporting (not finalising) documentation drafts

AI is not used to make autonomous regulatory decisions, approve strategies or determine safety conclusions.

Regulatory principle:

Human judgement must remain central, with clear accountability for decisions influenced by AI.

How we are aligned:

All platform outputs are:

• Reviewed by experienced scientific and regulatory experts
• Challenged where assumptions are weak or incomplete
• Validated before being used for execution, investment, or regulatory engagement

Responsibility for decisions always sits with identifiable humans, not the system.

Regulatory principle:

AI systems should be transparent, explainable, and interpretable in a regulatory context.

How we are aligned:

PreQlinic outputs are:

• Based on explicit rules, assumptions, and structured logic
• Traceable to defined inputs and regulatory drivers
• Explainable in terms familiar to regulators, investors, and domain experts

Outputs are designed to be interrogated, not accepted as opaque results.

Regulatory principle:

AI systems must rely on appropriate, relevant, and well-governed data sources.

How we are aligned:

PreQlinic does not rely on uncontrolled or indiscriminate data ingestion.

The platform:

• Uses curated, regulator-informed activity frameworks
• Applies deterministic rules rather than unconstrained pattern learning
• Separates data structuring from expert interpretation

This reduces the risk of spurious correlations or misleading optimisation.

Regulatory principle:

AI should be applied proportionately, with higher scrutiny where risk is greater.

How we are aligned:

AI is used specifically to:

• Surface uncertainty earlier
• Highlight areas of regulatory sensitivity
• Expose trade-offs before capital is committed

AI is not used where false precision or overconfidence could introduce risk.

Regulatory principle:

AI systems should be robust, reproducible and reliable over time.

How we are aligned:

PreQlinic is deployed through:

• Controlled validation programmes
• Iterative refinement based on expert review and real-world use
• Versioned logic and rule sets

Outputs are consistent, auditable, and reproducible within defined contexts.

Regulatory principle:

AI systems require ongoing governance across their lifecycle, not one-time validation.

How we are aligned:

AI components within PreQlinic are treated as governed infrastructure, including:

• Continuous monitoring of outputs
• Expert feedback loops
• Updates aligned with evolving regulatory expectations

AI is not treated as a static feature, but as a managed system.

Regulatory principle:

AI systems should not introduce or amplify bias, nor be used to circumvent regulatory scrutiny.

How we are aligned:

PreQlinic is explicitly designed not to:

• Optimise submissions to “game” regulatory review
• Replace regulatory reasoning with probabilistic scoring
• Mask uncertainty behind algorithmic outputs

The platform exists to strengthen regulatory discipline, not bypass it.

Regulatory principle:

AI-assisted processes should be auditable and appropriately documented.

How we are aligned:

PreQlinic outputs are:

• Documented with explicit assumptions and logic
• Structured for audit and review
• Suitable for investor, partner and regulatory scrutiny

This supports both internal governance and external review.

Regulatory principle:

AI frameworks should support consistency and collaboration across jurisdictions.

How we are aligned:

PreQlinic is designed to:

• Explicitly surface jurisdictional differences in regulatory expectations
• Support multi-region planning rather than implicit extrapolation
• Enable early comparison of pathways across regions

This reduces late-stage rework and regulatory misalignment.