At the AbedGraham Group
If you have a solution that includes AI capabilities, modules, or functionalities, it's likely your product will be considered a medical device and be subject to regulations such as:
- EU MDR (Medical Device Regulation)
- UK Medical Device Regulations
- NIS2 (cybersecurity directive)
At The AbedGraham Group, we provide expert advisory services for technology manufacturers building AI based products that are to be deployed within the UK and Europe. Our services support Class I self-certification through to higher risk product certification at Class IIa and above. Book a call or use our free tool to understand whether your product likely qualifies as a medical device.
Who This Service Is For
We work with:
- Digital health companies building or integrating AI features
- Medical device manufacturers
- Frontier AI labs
Whether you are preparing for Class I self-certification, or need to prepare for Class IIa/IIb/III requirements, our pragmatic approach to compliance ensures you meet requirements through a structured and time-efficient process.
We are specialists in compliance, cybersecurity and standards-based assurance for organisations operating in regulated healthcare markets.
AI Medical Device Compliance & Market Readiness
Class I AI Products: Self-Certification Support
Many health AI solutions fall under Class I due to the intended use claims and identified risks.
Pursuing Class I is a self-certification exercise in theory but in practice still requires significant document preparation before submission for certification.
We support manufacturers to:
- Define intended use
- Determine classification status
- Develop all relevant quality and technical documentation
- Complete risk management and usability analysis
- Baseline cybersecurity maturity
- Align with all relevant ISO standards
This service is optimised for manufacturers that need to move quickly to support market access but without taking risks with regulatory documentation.
Class IIa and Above
All AI products that impact clinical decision-making, diagnosis, or patient management will fall into Class IIa or higher.
At this classification level far more evidence, governance and operational readiness will be required and includes an external conformity assessment. We support organisations throughout the entire process apart from certification.
Core Areas of Support
ISO Standards & Quality Management
ISO standards are an essential part of compliance at this level. We work with clients to address these including but not limited to:
- ISO13485
- ISO14971
- ISO42001
- ISO27001
Cybersecurity & Resilience for AI Medical Devices
Cybersecurity is an essential part of AI medical device regulations and must be addressed in tandem with safety requirements.
As a leading healthcare cybersecurity consultancy firm we provide bespoke guidance addressing:
- Risk management
- Secure development
- Threat modelling and vulnerability management
- Business continuity and incident response
- Technical security documentation
Clinical Evaluation & Clinical Validation
Healthcare AI products are increasingly failing under regulatory scrutiny due to poor documentation relating to clinical safety and evaluation. As the UK's leading clinical safety specialists we support:
- Clinical risk analysis
- Clinical evidence gathering
- Clinical performance analysis
- Clinical safety documentation development
Post-Market Surveillance (PMS)
Post-market surveillance (PMS) has been enhanced under both UK and EU regulations.
We support:
- PMS documentation and process design
- PMS monitoring
- Incident handling
- Continuous improvement
AI products are subject to constant updates, retraining and model drift which must be monitored through stringent PMS systems.
Technical Documentation & Audit Readiness
We support manufacturers to draft and prepare all technical documentation subject to UKCA and CE marking requirements.
This includes:
- Quality management systems
- Risk management systems
- Cybersecurity evidence
- Clinical evidence
- Vigilance documentation
- Supplier assurance
AI-Specific Challenges
Healthcare AI products introduce additional complexity, including:
- Unclear or shifting intended use and claims
- Dataset provenance, bias and performance evidence
- Model drift and update governance
- Explainability and clinical interpretability
- Human factors and usability risk
- Safety and cybersecurity interdependence
We help you manage these challenges in a way that supports both innovation and regulatory confidence.
Why Choose The AbedGraham Group?
We are experts in clinical safety, cybersecurity and medical device regulations in the UK & Europe.
Our work supports:
- Faster market access
- Positive audit findings
- Reduced risk and liability
- Increased buyer confidence
Your business will scale successfully when supported by leading regulatory expertise.
AI & Medical Devices
Talk to an AI & Medical Devices specialist
If your AI product requires Class I self-certification support, or you need full compliance readiness for Class IIa and above, speak to our consultants today.
Frequently asked questions
Not necessarily. It depends on the intended use case, marketing claims and conditions/workflows being addressed. The greater the role an AI feature has in impacting clinical care, the more likely it will be classified as a medical device.
If your AI product meets the requirements of a Class I medical device then self-certification is the usual process. However, self-certification still requires all technical documentation and clinical validation to be completed first.
Being classed at IIa or above is usually due to the severity of conditions being addressed by a product and the level of intervention in clinical decisions being made by a solution or platform. For example, if AI is making a diagnosis in an emergency care scenario this will almost certainly be a Class III medical device.
Yes, cybersecurity standards and their implementation are a core part of medical device certification and post-market surveillance with a particular focus on secure development, vulnerability management and incident response.
This is a demonstration, using evidence, that a product behaves as per its intended clinical use case in a safe way. At Class I a literature review is often sufficient whereas at higher levels clinical data studies are often required.
AI products often require stronger post-market discipline because performance can change over time due to:
updates and retraining
changes in real-world data
workflow changes in clinical settings
model drift
Regulators expect manufacturers to monitor performance, manage change safely, and maintain continuous compliance through structured PMS processes.
Yes. We support AI-enabled medical device manufacturers across the full lifecycle - from Class I self-certification support through to full readiness for Class IIa and above, including ISO alignment, cybersecurity, clinical validation, technical documentation, and post-market surveillance systems.
Book a Discovery Call
If your AI product requires Class I self-certification support, or you need full compliance readiness for Class IIa and above, speak to our consultants today.