Ai In Autonomous Vehicle Patent Disputes.
AI IN AUTONOMOUS VEHICLE PATENT DISPUTES
Autonomous vehicles (AVs) rely heavily on AI for:
Self-driving systems (navigation, lane-keeping, collision avoidance)
Sensor data processing (LiDAR, radar, cameras)
Predictive control algorithms
Decision-making software for real-time traffic
This raises legal issues when AI creates or enables inventions, or when patents are infringed by AI-assisted designs.
Key Legal Issues:
Inventorship – Can AI be considered an inventor?
Ownership – Who owns AI-generated inventions?
Patentability – Are AI-generated autonomous vehicle designs patentable?
Infringement and liability – Who is liable when AI-assisted designs infringe patents?
1. UK Legal Framework
Patents Act 1977: Inventor must be a natural person.
Patentable subject matter: inventions must be new, inventive, and capable of industrial application.
AI cannot be listed as an inventor; human contribution is required.
Autonomous vehicle designs can be patented if they include technical solutions (e.g., sensor integration, control algorithms).
2. Key UK Case Laws
Case 1: Thaler v Comptroller General of Patents (UK, 2021–2023)
Facts:
Dr Stephen Thaler claimed that AI system DABUS invented autonomous vehicle systems.
Application listed AI as inventor.
Holding:
Only a natural person can be inventor.
Patent rejected; human intervention required.
Relevance:
AI cannot be an inventor for autonomous vehicle patents.
Human engineers or programmers who design or commission the AI must be listed as inventors.
Case 2: Autodesk Inc v Dyason (UK, 1992)
Facts:
Alleged infringement of CAD software outputs used in mechanical designs.
Human operator used software to generate outputs.
Holding:
Copyright/patent protection applies only with human intellectual input.
Software alone cannot hold rights.
Relevance:
AI used in autonomous vehicle design is treated as a tool, not an inventor.
Liability and patent rights rest with humans.
Case 3: Interlego AG v Tyco Industries (UK, 1989)
Facts:
CAD-generated LEGO brick designs were in dispute.
Holding:
Only original expression or inventive contribution by humans is protected.
Functional designs alone, even if optimized by a machine, do not automatically confer patent protection.
Autonomous Vehicle Implication:
AI-generated navigation algorithms may not be patentable without human inventive contribution.
Case 4: Lucasfilm Ltd v Ainsworth (UK, 2011)
Facts:
Replicas of Stormtrooper helmets; issue whether functional/technical aspects are copyrightable.
Holding:
Functional aspects are not protected under copyright.
Relevance:
For AV patents, AI-generated functional outputs must show technical invention to qualify.
Case 5: Sawkins v Hyperion Records Ltd (UK, 2005)
Facts:
Digital recording of historical music; human performer claimed rights.
Holding:
Human contribution protected; machine assistance irrelevant.
Implication for AI:
AI-assisted AV invention patents must have human intellectual contribution for legal protection.
3. US Legal Framework
35 U.S.C. §101–103: Defines patentable subject matter.
Inventor must be human (Thaler v USPTO).
Autonomous vehicle designs may be patented if they demonstrate novelty, non-obviousness, and technical implementation.
Case 6: Thaler v USPTO (2022)
Facts:
DABUS AI-generated inventions (vehicle control and safety systems).
Holding:
AI cannot be an inventor under US patent law.
Patent applications must identify human inventors.
Significance:
Strong precedent: AI cannot hold autonomous vehicle patents.
Case 7: American Axle & Manufacturing v Neapco Holdings (US, 2020)
Facts:
Patent claim on vibration damping system (mechanical control).
Holding:
Abstract mathematical ideas not patentable; technical application required.
Autonomous Vehicle Relevance:
AI-generated navigation or safety algorithms must demonstrate technical implementation, not just mathematical optimization.
Case 8: Alice Corp v CLS Bank (US, 2014)
Facts:
Patents related to abstract computer-implemented methods.
Holding:
Abstract ideas implemented on a computer are not patentable unless inventive concept exists.
Implication for AV AI:
AI-generated control methods must involve technical solutions beyond abstract logic.
Case 9: ClearPlay v Netflix (US, 2016)
Facts:
Automated content filtering technology.
Holding:
Liability arises for human developers/operators if automated tool infringes patents.
AI Relevance:
Human control is critical in AV patent enforcement when AI-generated design infringes.
4. EU Legal Framework
Directive 98/44/EC (Biotechnology & patents) – Inventor must be human.
Directive 2001/29/EC (Copyright) – Automated reproduction counts as infringement.
DSM Directive 2019/790 – Operators liable for AI-assisted infringement.
Case 10: CJEU – Infopaq v Danske Dagblades (C-5/08, 2009)
Facts:
Automated text extraction (analogous to AI outputs).
Holding:
Partial reproduction may infringe copyright.
Implication for AV AI:
AI systems reproducing patented designs can trigger operator liability.
Case 11: DABUS Patent Dispute in EU (EPO Appeals 2021)
Facts:
AI listed as inventor for vehicle-related patent applications.
Holding:
European Patent Office (EPO) rejected claims; inventor must be human.
Significance:
EU aligns with UK/US approach; human control critical in AV patents.
5. Comparative Analysis: UK vs US vs EU
| Aspect | UK | US | EU |
|---|---|---|---|
| Inventor | Human only | Human only | Human only |
| AI as inventor | ❌ | ❌ | ❌ |
| Technical requirement | Must have inventive contribution | Must demonstrate technical solution | Must demonstrate technical solution |
| Liability | Human user/operator | Human operator/developer | Human operator/developer |
| Patentable AV design | Yes, if technical | Yes, if technical | Yes, if technical |
| Abstract algorithms | ❌ | ❌ | ❌ |
6. Key Legal Principles
Human Inventorship – AI cannot be listed as inventor in any jurisdiction.
AI as a Tool – AI outputs are patentable only if human-guided creative contribution exists.
Technical Implementation Required – Mathematical or abstract optimizations alone are insufficient.
Operator/Developer Liability – Humans controlling AI systems can be held responsible for infringement.
International Consensus – UK, US, EU all reject AI as inventor; human intervention is mandatory.
7. Practical Implications for AV Companies
Document human input and decision-making in AI-assisted designs.
Ensure AI-generated inventions involve technical solutions, not just optimization.
File patents listing engineers, programmers, or system operators as inventors.
Maintain licensing agreements for AI software to minimize contributory liability.
Monitor AI outputs to prevent accidental infringement of third-party patents.
8. Summary Table of Cases
| Case | Year | Jurisdiction | Issue | Outcome | Relevance to AV AI Patents |
|---|---|---|---|---|---|
| Thaler v Comptroller | 2021–23 | UK | AI inventor | AI cannot be inventor | Human inventor required |
| Autodesk v Dyason | 1992 | UK | CAD outputs | Human input required | AI is a tool, not inventor |
| Interlego v Tyco | 1989 | UK | Functional designs | Originality needed | AI optimization insufficient |
| Lucasfilm v Ainsworth | 2011 | UK | Functional vs artistic | Functional not copyright | Technical design needed |
| Sawkins v Hyperion | 2005 | UK | Digital works | Human input protected | Human guidance required |
| Thaler v USPTO | 2022 | US | AI inventor | Patent denied | Human inventor required |
| American Axle v Neapco | 2020 | US | Abstract ideas | Not patentable | Technical contribution required |
| Alice Corp v CLS Bank | 2014 | US | Computer-implemented methods | Abstract ideas not patentable | Must be inventive technical solution |
| Infopaq v DD | 2009 | EU | Automated reproduction | Partial reproduction infringing | Operators liable for AI infringement |
| DABUS EPO | 2021 | EU | AI inventor | Rejected | Human inventor required |
This analysis shows that AI in autonomous vehicle patent disputes cannot hold inventorship, but its human operators, programmers, and designers bear full responsibility for inventorship and potential infringement.
RELATED Blog
comments