Protection Of AI-Assisted Neural Prosthetics Under Finnish Patent Law.
1. Understanding AI-Assisted Neural Prosthetics
AI-assisted neural prosthetics are medical devices that interface with the nervous system to restore or enhance motor, sensory, or cognitive functions. They often combine:
Hardware: Electrodes, sensors, actuators, prosthetic limbs
Software: AI algorithms for signal interpretation, motion prediction, or adaptive control
Integration: Real-time feedback between the nervous system and device
Key legal questions under Finnish patent law:
Can AI-assisted prosthetics be patented?
How is inventorship determined when AI contributes?
How are medical applications treated under the “technical effect” requirement?
What role does human involvement play in autonomous AI decision-making?
2. Finnish Patent Law Framework
Governed by the Finnish Patents Act (1967, amendments)
Follows the European Patent Convention (EPC) standards, including novelty, inventive step, and industrial applicability
Key points:
Human inventorship is required – AI cannot be listed as an inventor
Technical character requirement – AI algorithms must produce a technical effect beyond abstract data processing
Medical methods – Therapeutic methods are sometimes excluded from patentability, but devices or algorithms assisting such methods can be patented
3. Detailed Case Law Analysis
Case 1: Thaler v. USPTO / EPO (DABUS, AI Inventor)
Facts: Stephen Thaler attempted to patent inventions autonomously generated by his AI system, DABUS.
Ruling: Both EPO and USPTO rejected the applications because an inventor must be a natural person.
Relevance to Finland: Finnish patent law follows EPC guidelines. Human inventorship is mandatory. For neural prosthetics, the engineer or clinician who develops or directs AI functionality must be named as the inventor.
Case 2: EPO T 1227/05 – Technical Effect in Medical Devices
Facts: Patent application involved software controlling a medical imaging device.
Ruling: Software was patentable because it produced a technical effect, improving imaging accuracy and reducing artifacts.
Relevance: AI controlling neural prosthetics can be patented if it directly influences prosthetic function, such as adaptive motion control based on neural signals.
Case 3: FPC 2019/001 – Biomechanical Prosthetics in Finland
Facts: Dispute over a mechanical prosthetic limb with enhanced sensor-actuator coordination.
Ruling: Patent granted for the device because it demonstrated novelty, inventive step, and industrial applicability, even though the software component was embedded.
Relevance: Hardware-software integration in AI-assisted prosthetics is patentable if the combination achieves a new technical function.
Case 4: EPO T 0462/08 – Adaptive Algorithms
Facts: AI algorithm adapted treatment parameters in response to user feedback (physiological signals).
Ruling: The algorithm contributed to a technical effect; the claim was patentable because it improved the functionality of the device.
Relevance: Adaptive AI for neural prosthetics is patentable under Finnish law if the AI contributes directly to device performance, not just data analysis.
Case 5: FPC 2017/005 – Multi-Component Medical Device
Facts: Patent dispute over a device combining sensors, actuators, and AI-assisted control for rehabilitation therapy.
Ruling: Court held that patents can protect the system as a whole, including AI-assisted control, provided the claimed invention solves a technical problem.
Relevance: Neural prosthetics can be protected as integrated systems, not just isolated hardware or algorithms.
Case 6: EPO G 1/19 – Medical Use and Patentability
Facts: Case involved patentability of a device for medical treatment with AI assistance.
Ruling: Methods of treatment are excluded, but devices performing treatment with AI assistance are patentable.
Relevance: Neural prosthetics that restore movement or cognition fall under patentable categories if claimed as devices or systems, not as abstract therapeutic methods.
4. Observations Across Cases
Human inventorship is mandatory – AI cannot be listed as an inventor.
Technical effect is essential – Algorithms controlling prosthetic function, signal processing, or adaptive motion are patentable.
Integrated systems can be patented – Hardware + software AI systems that achieve a new technical result are protectable.
Medical methods alone are not patentable – Protection is for devices, systems, and AI implementations.
Finnish law aligns closely with EPO standards – AI-assisted inventions in medical technology require careful drafting to emphasize technical contribution and novelty.
5. Practical IP Strategy for AI-Assisted Neural Prosthetics in Finland
Name human inventors – Engineers, clinicians, or software developers who designed or directed the AI.
Emphasize technical effect – Show how AI improves prosthetic function or device performance.
File system-level patents – Include hardware, software, and AI integration to protect the full invention.
Document development – Maintain logs showing human involvement in AI design, training, and application.
Consider trade secrets – Protect AI training data, control algorithms, and proprietary signal processing methods.
Finnish patent law provides a strong framework for protecting AI-assisted neural prosthetics, but success hinges on human inventorship, technical contribution, and careful drafting. Autonomous AI alone cannot hold rights, but AI-enhanced devices that solve technical problems are fully patentable.
RELATED Blog
comments