The debate over whether AI systems can be considered inventors is just one example of the challenges ahead.
Artificial intelligence is transforming industries like healthcare, finance and technology. Yet protecting AI innovations, particularly algorithms, remains a legal challenge. This article explores the main legal obstacles to patenting AI algorithms and examines recent litigation that is shaping the future of intellectual property in this area.
Understand why algorithms are difficult to patent
Algorithms, which drive most AI systems by processing data and performing tasks, are often considered abstract ideas in patent law, making them difficult to protect. In the Alice Corp. case. v. CLS Bank International of 2014, the United States Supreme Court ruled that abstract ideas, such as algorithms or basic business practices, are not patentable unless they are applied in a concrete and innovative that goes beyond the abstract concept itself. .
The case arose when Alice Corporation attempted to patent a computerized method of managing financial transactions, and the Court ruled that the method was too abstract to qualify for a patent, thus setting a precedent for cases similar.
Since many algorithms are essentially mathematical instructions, they fall into this category. This makes it difficult for AI innovators to meet patent requirements for novelty, non-obviousness, and eligibility, limiting their ability to protect their inventions.
Patenting of AI algorithms in different regions
Legal standards for patenting AI innovations differ by region. In the United States, the Alice ruling led to the rejection of numerous AI patent applications. Algorithms often fail to meet the requirement of showing a concrete technological application, which frustrates many people in the field of AI development.
In Europe, the European Patent Office (EPO) has a more flexible standard. The EPO allows algorithms to be patented if they solve a technical problem or produce a “technical effect”. For example, an algorithm used in a self-driving car may be patentable, whereas a more general-purpose AI system might not be. This approach makes Europe a more favorable place for AI innovators seeking patents.
China, on the other hand, has been even more willing to grant AI patents. As part of its broader strategy to become a leader in AI, the Chinese Patent Office is becoming more flexible on AI-related applications, allowing patents that might not be granted in the United States or Europe. This variation in global patent standards creates uncertainty for AI innovators trying to protect their work in different markets.
Recent legal disputes shape AI patent law
Several legal cases highlight the ongoing challenges between AI innovation and current intellectual property laws. One of the key cases is Thaler v. Controller General of Patents, involving an AI system called DABUS, created by Stephen Thaler.
Thaler attempted to list DABUS as the inventor in patent applications filed in various countries, including the United States and the United Kingdom. Courts have ruled that only human inventors can be listed, rejecting the idea that AI can be considered an inventor. This case has sparked debate about the role of AI in the patent system and whether patent laws should adapt to recognize non-human inventors.
Strategies to protect AI innovations
Given the challenges of patenting algorithms, AI innovators should explore other strategies to protect their work. One approach is to focus on patenting specific applications of an algorithm rather than the algorithm itself. If an algorithm solves a technical problem in an area like healthcare or autonomous vehicles, it is more likely to be patented.
Another option is to rely on trade secrets. Many companies choose to keep their AI algorithms confidential, using non-disclosure agreements and internal security measures. Although trade secrets do not provide the same level of legal protection as patents, they can be effective, particularly for algorithms that are difficult to reproduce.
A hybrid strategy can also work. Some companies patent specific applications while keeping the underlying algorithms secret. This allows them to obtain legal protection without revealing the core technology behind their innovation.
The Future of AI Patent Law
As AI technology develops, patent law will need to evolve. The debate over whether AI systems can be considered inventors is just one example of the challenges ahead. Additionally, global competition could lead to legal changes as specific regions seek to attract AI innovators.
For now, the challenge remains: while AI is the engine of technological progress, its algorithms are difficult to patent under current laws. AI innovators should stay informed about the legal landscape and consider other ways to protect their technologies.
