Ai Diagnostic Tool Ip Disputes Uk.
AI Diagnostic Tools and IP Disputes in the UK
AI diagnostic tools in healthcare are software systems or platforms that analyze patient data to assist clinicians in diagnosis, risk prediction, or treatment recommendations. These tools often involve complex algorithms, machine learning models, and proprietary datasets, raising multiple IP and legal issues in the UK.
IP disputes generally arise in the following areas:
Patent Infringement – patented algorithms, AI methods, or AI-augmented medical devices.
Copyright – software code, training datasets, or user interfaces.
Trade Secrets – proprietary algorithms, training datasets, or model parameters.
Regulatory/Clinical Compliance Conflicts – indirectly impacting IP enforcement when tools are used without proper licensing.
1. Patent Disputes Involving AI Diagnostic Tools
Case 1: Fisher & Paykel Healthcare Ltd v ResMed Ltd (2007)
Facts: Dispute over patents related to medical devices for respiratory care.
Issue: Whether ResMed’s device infringed Fisher & Paykel’s patents.
Judgment:
Court emphasized that patent claims must be clearly defined and novel.
Infringement occurs if a competing device falls within the scope of patent claims.
Relevance to AI: Patent disputes for AI diagnostic tools follow similar principles: the algorithm, method, or system must be clearly claimed, and unauthorized replication constitutes infringement.
Case 2: Halliburton Energy Services v Smith International (2006)
Facts: Software and method patents for diagnostic technology were disputed.
Issue: Whether software-implemented methods were patentable.
Judgment:
UK courts recognized that computer-implemented inventions may be patentable if they produce a technical effect.
Pure algorithms without a technical effect are not patentable.
Relevance: AI diagnostic tools may be patentable if they demonstrate improvement in medical diagnosis, patient safety, or clinical workflow.
Case 3: Actavis v Eli Lilly (2015)
Facts: Patent infringement involving pharmaceutical methods, but principles apply to AI diagnostic tools.
Issue: Whether patent claims are infringed by similar technology.
Judgment:
Court allowed a broad interpretation of claims, covering similar methods that achieve the same technical effect.
Relevance: AI algorithms performing equivalent diagnostic steps may infringe patents even if the implementation differs.
2. Copyright Disputes
AI diagnostic tools rely heavily on proprietary software and datasets. Copyright protection applies to:
Software code
User interfaces
Curated datasets
Case 4: SAS Institute Inc v World Programming Ltd (2013)
Facts: WPL replicated the functionality of SAS software without copying source code.
Issue: Whether replicating functionality infringed copyright.
Judgment:
Copying functionality is not copyright infringement; only the code or expression is protected.
Relevance to AI: Competitors cannot copy AI code, but developing a new AI tool with the same functionality using original code is lawful.
Case 5: Navis v Softship (2014)
Facts: AI-style predictive software for maritime logistics. Copyright dispute arose over user interface and code.
Judgment:
Copyright protection extended to original structure and expression, not underlying ideas.
Relevance: AI diagnostic software’s code, UI design, or data pipelines may be protected, but the medical ideas behind the algorithm cannot.
3. Trade Secret / Confidential Information Disputes
AI diagnostic tools often rely on proprietary datasets, model weights, or hyperparameters.
Case 6: Kerry Ingredients v Bakkavor (2016)
Facts: Bakkavor misused confidential technical information about food products.
Judgment:
Misuse of confidential information giving a competitive advantage is actionable.
Relevance to AI: Using proprietary AI training data or model parameters without authorization constitutes trade secret misappropriation.
Case 7: IBM v LzLabs (2025)
Facts: Reverse engineering of proprietary software was disputed.
Judgment:
Misuse of confidential knowledge, even if reimplemented differently, breached trade secret principles.
Relevance: Extracting or replicating proprietary AI models from diagnostic tools can trigger liability under UK trade secret law.
4. Key Principles Emerging from UK Case Law
| Principle | Explanation | AI Application |
|---|---|---|
| Patent protection | Computer-implemented methods with technical effect can be patented | AI diagnostic algorithms that improve diagnosis accuracy are patentable |
| Copyright | Protects source code and UI, not underlying ideas | Code replication is prohibited, but functional equivalence is allowed |
| Trade secrets | Misuse of confidential information is actionable | Proprietary datasets, hyperparameters, and model architectures are protected |
| Reverse engineering | Allowed if independent effort, not through misuse | Lawful reimplementation is allowed; stealing training data or model info is not |
| Scope of claims | Courts may interpret claims broadly | Similar AI tools performing equivalent methods may infringe |
5. Practical Implications for AI Diagnostic Tools
Patents: Developers should ensure algorithms producing clinical benefits are patentable and avoid infringing existing AI or medical patents.
Copyright: Maintain original code; avoid copying competitors’ software.
Trade Secrets: Safeguard datasets, model weights, and hyperparameters with NDAs, access controls, and encryption.
Regulatory Compliance: Misuse of AI diagnostic tools may trigger both IP and healthcare regulation issues.
6. Conclusion
In the UK, IP disputes over AI diagnostic tools involve patents, copyright, and trade secrets, often intersecting with software law and healthcare regulation. Key lessons from case law:
Fisher & Paykel v ResMed – patent enforcement for medical devices.
Halliburton v Smith – patentability requires technical effect.
SAS v WPL – copyright protects code, not ideas.
Kerry Ingredients v Bakkavor – trade secrets are actionable.
IBM v LzLabs – misappropriation applies even if reimplemented differently.
UK courts balance innovation with protection, ensuring AI diagnostic tools are safeguarded while encouraging independent development.
RELATED Blog
comments