Ai Intellectual Property Dispute Resolution Mechanisms.
AI INTELLECTUAL PROPERTY DISPUTE RESOLUTION
Overview
AI is increasingly involved in IP, raising disputes in:
Copyright – AI-generated content, training datasets
Patents – AI inventions, AI-assisted prior art
Trade secrets – AI models, proprietary algorithms
Trademark – AI-generated brand or design confusion
Dispute resolution mechanisms include:
Judicial litigation (courts)
Alternative Dispute Resolution (ADR): arbitration, mediation
Regulatory review: IP offices (USPTO, EPO, WIPO)
The challenge: AI blurs traditional IP boundaries, especially around authorship, inventorship, ownership, and liability.
I. Key Case Laws and Their Implications
CASE 1: Thaler v. Vidal (US, 2020–2022)
Facts:
Stephen Thaler named his AI system “DABUS” as inventor on US patent applications.
Dispute:
AI inventor recognition under patent law.
Outcome:
USPTO rejected; only natural persons can be inventors.
Thaler appealed, but courts confirmed rejection.
Mechanism Used:
Judicial litigation (USPTO and Federal courts)
Principle:
Dispute over inventorship cannot assign AI legal personhood; human involvement is required.
Relevance for IP Dispute Resolution:
AI-assisted inventions require human claimants for judicial resolution.
Patent offices are the first point of dispute; courts finalize.
CASE 2: European Patent Office DABUS Applications (2021–2023)
Facts:
DABUS inventor patent applications filed in Europe.
Dispute:
AI recognition as inventor under European Patent Convention (EPC).
Outcome:
EPO rejected; inventor must be a person.
Appeal pending, but consistent with USPTO.
Mechanism:
Administrative review (EPO Boards of Appeal)
Could escalate to judicial review in national courts.
Principle:
Administrative IP bodies serve as primary dispute resolution for AI inventorship issues.
CASE 3: Midler v. Ford Motor Co. (1988)
Facts:
Ford used a voice impersonator to imitate Bette Midler.
Dispute:
Right of publicity / copyright in distinctive human voice.
Outcome:
Court ruled in favor of Midler; imitation of voice for commercial advantage violated rights.
Mechanism:
Civil litigation
Principle:
For AI-generated TTS voices, courts will treat recognizable human likeness or voice as actionable IP violation.
ADR or licensing could resolve disputes preemptively.
CASE 4: Waits v. Frito-Lay (1992)
Facts:
Frito-Lay hired a sound-alike singer for a commercial.
Dispute:
Right of publicity / false endorsement.
Outcome:
Waits won; imitation created market confusion and violated publicity rights.
Mechanism:
Judicial litigation
Principle:
AI-generated imitation of celebrity voices or styles can trigger IP disputes; ADR like mediation can limit reputational or financial damage.
CASE 5: Authors Guild v. Google (2015)
Facts:
Google scanned millions of books and displayed snippets.
Dispute:
Copyright infringement; licensing of digital content.
Outcome:
Court held it was fair use because transformative.
Mechanism:
Judicial litigation, with settlement negotiations
Principle:
AI training disputes often resemble traditional copyright cases; licensing disputes can be resolved through court or negotiation.
AI companies may preempt disputes via license agreements and ADR clauses.
CASE 6: Andersen v. Stability AI (2023–Ongoing)
Facts:
Artists sued AI companies for training on copyrighted images without consent.
Dispute:
Copyright infringement, derivative works.
Mechanism:
Litigation and potential class-action settlement
Early negotiation attempts reported
Principle:
AI-generated outputs may create derivative work disputes; ADR mechanisms like arbitration can help speed resolution and manage risk.
CASE 7: Getty Images v. Stability AI (2023–Ongoing)
Facts:
Getty Images claimed AI used its licensed images for model training; outputs contained watermarks.
Dispute:
Unauthorized use of licensed content; licensing disputes.
Mechanism:
Litigation
Possible settlement negotiations as risk mitigation
Principle:
Traceable infringement increases likelihood of resolution via settlement or arbitration before court trial.
CASE 8: TikTok Text-to-Speech Voice Litigation (2023)
Facts:
Voice actor claimed TikTok used her recordings for AI voice TTS system without consent.
Dispute:
Right of publicity and copyright.
Mechanism:
Litigation filed in US courts
Early negotiations indicated interest in ADR mediation
Principle:
For AI voice disputes, companies often pursue out-of-court settlements to avoid injunctions or reputational harm.
II. Mechanisms for AI IP Dispute Resolution
1. Judicial Litigation
Courts resolve disputes over:
AI inventorship
Copyright ownership
Trade secret misappropriation
Pros: legally binding, authoritative
Cons: slow, expensive, may set binding precedent
Relevant Cases: Thaler v. Vidal, Midler v. Ford, Andersen v. Stability AI
2. Administrative and Regulatory Review
Patent offices, copyright offices, and IP boards handle disputes in first instance:
AI patent filings
Licensing disputes
Trademark opposition for AI-generated brands
Relevant Cases: EPO DABUS applications
3. Alternative Dispute Resolution (ADR)
Arbitration and mediation are increasingly used for AI IP disputes:
Licensing disagreements
Training dataset use
AI-generated outputs
Pros: faster, confidential, preserves business relationships
Cons: no precedent-setting, outcome depends on agreement
Relevant Cases: Early stages in Andersen v. Stability AI and TikTok TTS litigation
4. Hybrid Approaches
Pre-litigation negotiation + ADR + limited court enforcement
Common for AI licensing agreements and TTS voice disputes
III. Key Principles for AI IP Dispute Resolution
AI cannot be an inventor; human oversight required (Thaler, DABUS)
Transformative use may reduce copyright liability, but derivative outputs are risky (Authors Guild v. Google, Andersen)
Recognizable human elements (voice, likeness) are protected (Midler, Waits)
Licensing agreements and contractual clauses are critical
ADR is preferred for rapid, confidential resolution, especially for commercial AI applications
Administrative bodies provide first-instance relief before courts
IV. Recommendations for AI IP Dispute Management
Include IP clauses in AI training dataset agreements
Ensure consent and licensing for human voices, artwork, or text
Maintain audit trails for AI outputs to reduce litigation risk
Use mediation or arbitration for international disputes
Engage patent/copyright offices early for AI-assisted invention filings
V. Summary
AI IP disputes are multi-dimensional:
| Aspect | Case Example | Mechanism |
|---|---|---|
| Inventorship | Thaler v. Vidal | Courts |
| AI-assisted patent validity | EPO DABUS | Regulatory |
| Copyright infringement | Authors Guild v. Google | Litigation & ADR |
| Voice imitation | Midler v. Ford | Courts & ADR |
| Dataset licensing | Getty Images v. Stability AI | Litigation/Settlement |
| TTS voice rights | TikTok TTS | Litigation/ADR |
| AI-generated derivative works | Andersen v. Stability AI | Litigation/ADR |
Key takeaway: AI IP disputes are increasingly resolved via a mix of regulatory review, litigation, and ADR, with a strong emphasis on human oversight, licensing, and consent.
RELATED Blog
comments