The UK Medicines and Healthcare products Regulatory Agency has placed artificial intelligence at the centre of its five-year regulatory strategy.
Published on 14 October 2026, the MHRA Strategy 2030 describes how the Agency intends to modernise medicines and medical-device regulation in response to rapid advances in artificial intelligence, data science, personalised medicines and advanced therapies.
The significance of the strategy is not simply that MHRA supports innovation.
It contains several concrete AI commitments, including:
- development of an AI-enabled Clinical Trial Protocol Checker;
- publication of an AI Medicines Development Action Plan in 2027;
- establishment of an MHRA Centre for AI and Digital Health;
- use of AI and real-world data to identify safety issues earlier;
- movement toward continuous, real-time safety assurance.
For pharmaceutical companies, this is an important regulatory signal:
AI is moving from isolated pilot projects into the core infrastructure of medicines development, regulatory assessment and post-market oversight.
AI Challenges Traditional Regulatory Models
MHRA states that artificial intelligence is transforming both regulated products and the way organisations operate.
The Agency also recognises that these developments challenge traditional regulatory models and require new approaches to:
- evidence generation;
- clinical research design;
- manufacturing oversight;
- post-market safety;
- surveillance.
This is particularly important for pharmaceutical companies.
AI is no longer relevant only where it is incorporated into a digital medical device.
It can also affect:
- drug discovery;
- nonclinical modelling;
- clinical-trial design;
- regulatory submissions;
- manufacturing and quality systems;
- pharmacovigilance;
- regulatory review.
The same organisation may therefore use AI in several regulated contexts, each requiring a different level of assurance.
A New AI Medicines Development Action Plan Is Coming in 2027
One of the most important commitments is the planned publication of an:
AI Medicines Development Action Plan in 2027.
The strategy announcement states that the plan will explain how AI will be used to transform medicines development.
This could become an important UK regulatory reference for pharmaceutical companies.
Key questions that the future Action Plan may need to address include:
- how AI Context of Use should be defined;
- what evidence is required to demonstrate model credibility;
- how training and validation data should be governed;
- how AI-supported evidence should be documented;
- how model changes should be controlled;
- what level of human oversight is expected;
- how performance should be monitored throughout the lifecycle.
The Action Plan has not yet been published, and its detailed scope remains to be seen.
Nevertheless, committing to a dedicated medicines-development plan shows that MHRA intends to move beyond broad responsible-AI principles toward application-specific regulatory expectations.
An AI-Enabled Clinical Trial Protocol Checker
MHRA also plans to develop an AI-enabled Clinical Trial Protocol Checker.
Its intended purpose is to provide sponsors with greater regulatory certainty and support faster, higher-quality clinical-trial applications.
This is a significant example of AI operating on the regulator side of the regulatory interface.
A protocol checker could potentially help identify:
- missing information;
- internal inconsistencies;
- common application deficiencies;
- potential regulatory issues;
- areas requiring clarification before submission.
The objective would not necessarily be to replace regulatory assessment.
Instead, AI could help sponsors improve submission quality before formal review.
AI Will Increasingly Operate on Both Sides of the Submission
Pharmaceutical companies are already considering AI for:
- protocol drafting;
- regulatory writing;
- data analysis;
- quality control of submissions;
- regulatory intelligence.
At the same time, regulators are adopting AI to support:
- application triage;
- protocol assessment;
- information retrieval;
- evidence analysis;
- safety surveillance.
This creates a new regulatory environment:
Sponsor AI → Regulatory Submission → Regulator AI → Human Regulatory Decision
As AI becomes more common on both sides, the quality and traceability of the underlying information may become increasingly important.
Regulatory submissions may need to become more:
- structured;
- machine-readable;
- internally consistent;
- traceable to authoritative sources;
- suitable for automated analysis.
A New MHRA Centre for AI and Digital Health
The strategy also commits MHRA to establishing a:
Centre for AI and Digital Health.
The Centre is intended to clarify regulatory expectations for developers of digital and AI-driven medical devices and medicines.
This could help address a recurring industry problem.
Organisations often understand the high-level AI principles but remain uncertain about practical implementation.
Questions may include:
- Which regulatory framework applies?
- What validation evidence is expected?
- How should adaptive AI be controlled?
- How should third-party foundation models be governed?
- What information should be included in a regulatory submission?
- How should post-deployment performance be monitored?
A dedicated centre could provide a more coherent regulatory interface for these questions.
From Periodic Safety Review to Continuous Safety Assurance
Another major commitment concerns post-market surveillance.
MHRA intends to move toward:
continuous, real-time safety assurance using data, analytics and AI.
The Agency plans to:
- improve access to hospital and genomic data;
- use AI to analyse real-world data;
- identify potential safety issues earlier;
- work with the Health Data Research Service;
- strengthen real-time safety-signal detection;
- simplify Yellow Card adverse-event reporting.
This reflects a broader shift in pharmacovigilance.
The traditional model relies heavily on periodic reporting and retrospective analysis.
AI-supported surveillance could increasingly provide:
- earlier signal identification;
- automated case clustering;
- trend detection;
- analysis across multiple data sources;
- continuous reassessment of product safety.
AI Does Not Remove the Need for Scientific Oversight
More continuous surveillance does not mean that every statistical or algorithmic signal represents a genuine safety problem.
AI may identify correlations, unusual patterns or clusters that require further evaluation.
Qualified experts must still determine:
- whether the signal is scientifically plausible;
- whether alternative explanations exist;
- whether data quality is sufficient;
- whether regulatory action is necessary.
A mature safety architecture therefore remains:
AI Signal Detection → Scientific Assessment → Human Decision → Regulatory Action
Implications for Manufacturing and GMP
MHRA Strategy 2030 is not a new GMP guideline.
However, it explicitly states that technological developments require innovative approaches to manufacturing oversight.
This matters because AI is increasingly being considered for:
- process monitoring;
- predictive maintenance;
- deviation analysis;
- quality trending;
- laboratory analytics;
- advanced process control;
- manufacturing knowledge management.
The strategy does not yet specify how MHRA will evaluate these individual use cases.
Nevertheless, it suggests that manufacturing oversight will need to evolve as digital and AI-enabled systems become more common.
For pharmaceutical manufacturers, this supports the need to establish AI governance before specific inspection expectations become fully developed.
The Regulator Is Also Becoming an AI Organisation
The strategy is not only about regulating AI products.
It is also about using AI to modernise MHRA’s own operations.
This distinction is important.
The Agency is simultaneously becoming:
- a regulator of AI;
- a user of AI;
- a producer of AI-supported regulatory analysis.
This may influence how regulatory information is submitted, reviewed and monitored.
It may also create new expectations for data quality, consistency and digital interoperability across the medicines lifecycle.
What Pharmaceutical Companies Should Do Now
The publication of the strategy does not create an immediate new compliance obligation.
However, companies can begin preparing for the regulatory direction it describes.
Useful actions include:
- maintaining an inventory of AI use cases;
- defining intended use and Context of Use;
- classifying GxP and regulatory impact;
- documenting data provenance;
- establishing human-accountability requirements;
- controlling models, configurations and suppliers;
- monitoring operational performance;
- following development of the 2027 AI Medicines Development Action Plan.
Companies operating in the UK should also consider how AI-supported regulatory submissions may interact with future MHRA tools such as the Clinical Trial Protocol Checker.
Why This Strategy Matters
MHRA Strategy 2030 represents more than a general statement that artificial intelligence is important.
It establishes several concrete deliverables:
AI Protocol Checker → AI Medicines Development Action Plan → AI Centre → AI-Enabled Safety Surveillance
Together, these commitments show that AI is becoming part of the regulator’s long-term operating model.
The relevant question for pharmaceutical companies is therefore shifting from:
“Will regulators eventually address AI?”
toward:
“How should our systems, data, governance and workforce prepare for an AI-enabled regulatory environment?”
Regulatory Status Note
MHRA Strategy 2030 is a strategic policy document.
It does not itself establish new legally binding GMP requirements or detailed validation expectations for pharmaceutical AI.
The AI Medicines Development Action Plan is planned for 2027, and the Clinical Trial Protocol Checker and Centre for AI and Digital Health remain future commitments.
The strategy should therefore be understood as a strong regulatory-direction signal rather than a current AI compliance standard.
Sources
MHRA – MHRA Strategy 2030:
https://www.gov.uk/g … s/mhra-strategy-2030
MHRA – Strategy 2030 executive summary:
https://www.gov.uk/g … 30-executive-summary
MHRA – Strategy 2030 announcement: