Artificial Intelligence and Inventorship Under the Patent Law Regime: Practical Developments from Common Law Jurisdictions

The emergence of Artificial Intelligence (AI) has sparked a growing debate around its role in innovation and inventorship, challenging traditional patent law structures. Given the capacity of AI to autonomously generate solutions, designs, and even inventions, many common law jurisdictions face unprecedented questions regarding the attribution of inventorship and intellectual property (IP) rights. This article explores the current legal landscape, practical implications, and recent developments in AI-related patent law across common law jurisdictions, focusing on the United States, United Kingdom, and Australia.

The AI Dilemma in Patent Law

Under traditional patent law, inventorship is strictly reserved for human beings. Inventors are required to meet specific criteria, including conception and reduction to practice, which require elements of mental input, creativity, and intention. However, as AI technologies, particularly machine learning and neural networks, evolve to generate innovative solutions autonomously, they challenge the notion that only human beings can "conceive" inventions. Such AI-generated inventions raise questions as to whether AI systems can or should be named as inventors, who should hold the rights to such patents, and how these rights should be distributed.

The U.S. Stance on AI as Inventors

In the United States, the patent law landscape is largely shaped by the U.S. Patent Act and common law precedents. The U.S. Patent and Trademark Office (USPTO) and U.S. courts have consistently held that only natural persons can be recognized as inventors. A recent landmark case, Thaler v. Hirshfeld, addressed the inventorship of an AI-created invention for the first time. The case revolved around Stephen Thaler, who submitted a patent application for an invention created by his AI system, known as DABUS. The USPTO rejected the application, affirming that U.S. law requires a human inventor, and the federal court upheld this decision, reinforcing the interpretation that AI cannot be considered an "inventor" under the Patent Act.

This strict stance reflects a traditional view on inventorship, which emphasizes human creativity and intentionality, underscoring that AI lacks the legal personhood necessary for patent rights. However, the ruling has led to questions about who holds rights over AI-generated inventions if AI cannot legally be an inventor. The law in the U.S. thus continues to grapple with recognizing inventorship in a way that both promotes innovation and aligns with current legal frameworks.

The United Kingdom’s Approach to AI-Generated Inventions

The UK, another common law jurisdiction, has similarly faced challenges in recognizing AI as inventors under its patent laws. In 2021, the UK Court of Appeal heard the case of Thaler v. Comptroller General of Patents, which also involved DABUS-generated inventions. Like in the U.S., the court rejected the application on the grounds that an inventor must be a human being and not an AI. The court's decision was based on a similar interpretation of the Patent Act, requiring a “natural person” as the inventor.

Despite rejecting AI as an inventor, UK courts have acknowledged the complexities of AI’s role in innovation and have called for more nuanced discussions in legislative and policy circles. The UK Intellectual Property Office (IPO) has since undertaken consultations on AI and IP, aiming to explore alternative frameworks for handling AI-generated innovations, potentially shifting towards a more flexible approach. However, as of now, UK patent law remains firmly rooted in human-centered inventorship.

Australia’s Emerging Perspective on AI Inventorship

In a surprising divergence from other common law jurisdictions, Australia’s Federal Court ruled in 2021 that an AI system could be considered an inventor. In Thaler v. Commissioner of Patents, the Australian court found that under the Australian Patents Act, there was no requirement that an inventor must be a natural person. This groundbreaking decision recognized DABUS as the inventor, marking a significant departure from established international norms.

The court's ruling was primarily based on the belief that patent law should serve as a tool for incentivizing innovation, regardless of whether the inventor is human. However, the Australian High Court overturned the decision in 2022, aligning more closely with the global consensus that only natural persons can be inventors. Nevertheless, this brief allowance of AI as an inventor in Australia has spurred discussions on whether patent laws should be adapted to accommodate non-human inventors.

Practical Implications of AI in the Patent System

1. Ownership and Rights Allocation: A key question arising from AI-driven inventions is the allocation of rights. Without a clear inventor, there are challenges in determining who holds the patent rights. If the developer of an AI system is considered the de facto inventor, it may create an environment where AI developers monopolize patents, potentially stifling broader innovation.

2. Legal Certainty and Economic Incentives: Uncertainty around AI-related patents could discourage companies from investing in AI research. Ensuring legal certainty around AI-driven innovations is crucial to foster economic growth, innovation, and competitiveness. Jurisdictions that adapt their laws to recognize AI's contributions could become leaders in attracting tech companies and innovators.

3. Ethical and Moral Considerations: AI inventorship raises ethical questions about the nature of creativity and whether a machine, lacking consciousness, can "invent" in a meaningful way. Granting rights to non-humans challenges traditional views of autonomy, personhood, and responsibility, leading to calls for legislation that considers these philosophical dimensions.

4. International Harmonization of AI Patent Laws: Different stances on AI inventorship across jurisdictions could lead to legal fragmentation, complicating patent applications, enforcement, and cross-border IP protection. International bodies, such as the World Intellectual Property Organization (WIPO), are exploring harmonized frameworks to address these challenges and ensure a consistent approach to AI-related patents.

   
   

Looking Ahead: Potential Reforms and Future Directions

As AI continues to advance, legislative bodies worldwide are under pressure to revisit patent laws. The practical implications of AI-generated inventions extend far beyond inventorship alone, potentially reshaping innovation, competition, and international patent strategies. Moving forward, common law jurisdictions might consider several reforms, including:

• Amendments to Inventorship Criteria: Revising patent laws to clarify the definition of “inventor,” potentially including a “human contribution” requirement, could provide a balanced approach to AI-related inventorship.

• Special Designations for AI-Generated Inventions: Creating a separate legal designation for AI-generated innovations could preserve human inventorship requirements while recognizing AI’s contributions in the innovation process.

• Rights Allocations to AI Developers and Users: Legislators could explore assigning rights to either the developers or users of AI systems, based on the level of control and direction exercised over the invention process.

  
  

Conclusion

The debate over AI and inventorship highlights the challenge of reconciling centuries-old patent laws with rapid technological progress. Common law jurisdictions like the U.S., UK, and Australia continue to navigate this evolving legal landscape, weighing the potential for innovation against the integrity of established IP frameworks. As discussions progress, there may be opportunities for reforms that balance the protection of human creativity with the recognition of AI’s role in driving the next wave of invention. Addressing these challenges will be crucial for a cohesive and adaptable patent system capable of fostering innovation in the age of artificial intelligence.