IP law presents several challenges for AI companies that develop and commercialize intelligent systems, including patent eligibility. Patent eligibility in AI contexts is challenging, even for functional claims. Embodiments must be disclosed, however, so that patents can cover the invention. In addition, AI systems are often difficult to classify as “invented” in the current IP landscape. In this article, we will discuss some aspects of AI patent eligibility and how AI systems can be protected by IP law.
Trade secret protection
Whether or not trade secret protection is right for your AI invention depends on several factors. Patent protection can be expensive, uncertain, and not always applicable to the subject matter of the invention. For example, many software patents are invalidated under the Alice decision, which made them publicly available. If you fail to keep your AI invention confidential, you risk losing the benefit of patent protection altogether. In such cases, trade secret protection may be more appropriate.
Another reason to protect AI trade secrets is to prevent reverse engineering. Often, competitors will attempt to reverse engineer new technology, which can erode trade secret protection. While it may be difficult to reverse-engineer complex AI today, it may be easier in the future. In these cases, trade secret protection for AI is almost redundant. So, how can you ensure that your AI is protected? First, it is important to follow a few basic rules.
Patent rights can protect the most obvious features of an AI/ML platform, but trade secrets can cover a broad range of information. They may include proprietary training data, formulas used in the model, and the way in which the output of the AI/ML platform is used. Other trade secrets might include aspects of the platform itself. These can all be vulnerable to compromise. If you can protect your trade secrets, you can ensure that your company’s innovation is protected.
Second, you need to take appropriate measures to prevent traces of your trade secrets from getting out. For example, you must require all persons who have contact with the confidential information to sign a confidentiality agreement. You should also insist that any agreements contain specific terms related to the content of trade secret protection. A version control system for operational software can be useful in identifying who owns the rights. You can even choose the most similar version for comparison purposes.
Lastly, copyright protection should be considered as a part of your AI/ML protection strategy. It provides supplementary protection for certain portions of an AI system and also helps preserve your rights in the event of infringement. Copyright protection is essential for protecting your AI/ML platform, and you should consider it alongside trade secret protection. So what can trade secret protection for artificial intelligence do for you? Find out more about how trade secret protection can help you protect your work today.
Copyright protection
In recent years, there have been numerous reports highlighting the need for copyright protection for AI technologies. However, there is one problem. The legal status of AI technologies cannot be determined without considering human authorship. The EU’s Economic and Social Committee has rejected the idea of giving AI the legal status of author. However, the CLA has pointed to the necessity of reviewing the originality standard. Even so, the CLA is not alone in its concern.
The Whelan court incorrectly relies on the earlier Synercom decision and holds that input formats of computer programs are more specific than functional design and line-by-line program design. Despite the ruling, the Supreme Court has refused to consider the Whelan decision. Meanwhile, the European Commission introduces a sui generis layout protection system, modeled on the US Semiconductor Chip Protection Act. Moreover, the Japanese district court wrongly concludes that the arrangement of algorithms in a foreign language is protectible.
The Japanese government has proposed bypassing copyright protection by creating new types of rights for creators of AI. These new rights could be similar to the database right that protects intellectual works in Europe. Such a right would protect investors’ investment in AI. But to date, no such legislation has been enacted. In the meantime, the world should be aware of the growing risks of copyright protection for artificial intelligence. So, how can we protect ourselves against this emerging trend?
The United Kingdom has passed a law granting copyright protection to AI in 1988. This decision recognizes the importance of ensuring that AI forms have equal rights with human creators. It is also important to note that the United States prohibits copyright protection for AI creations unless they are created by humans. The law is changing, but it is a step in the right direction. This case is a good example of how AI can be protected.
In its Final Report, the CONTU panel had dedicated 58 pages to software copyright. The opinion in the report supported software copyright protection while the 28-page report was against it, despite the opinions of a few members of the commission who had some knowledge of computing. This confusion eventually spilled into the 1982 court case. In this case, the two arguing counsels, Miller and Levine, claimed that copyright protection does not apply to software.
Patent protection
Some researchers believe that AI systems are more inventive than humans and that their ideas should be given patent protection. However, the idea of patent protection for AI systems is not entirely without danger. While AI systems could speed up the creation process, they could also overburden the patent system and widen the gap between the haves and the have-nots. Moreover, AI systems could also change the nature of an invention. They could be more knowledgeable than a person skilled in the art.
Moreover, AI inventors could use the patent to distribute their creations. In this way, AI-generated inventions could benefit the developer of the AI tool, the person directing it, or the owner of the data that was used to train the system. This approach has many benefits, including allowing the invention to be widely distributed. AI tools may be the future of artificial intelligence and they should receive patent protection as soon as possible.
Moreover, it is important to note that patents will only be granted if they are tied to a specific real-world problem. This is a requirement in many jurisdictions, including the European Patent Office. In the past, claiming an AI invention did not require a person of ordinary skill in the art, but it does require that the invention be applicable in a specific field of technology. Meanwhile, patent offices have taken a technical approach and treat AI elements as software.
While AI inventorship may be simpler than the equivalent regime for human inventors, there are some concerns. Some researchers believe that it is possible for employees to claim patent protection for an AI invention if they coded or operated the machine in the course of their employment. While appeals against such decisions are unlikely, the global patent office’s decision confirms that AI inventors must be natural persons. If they do, it is essential that they have human rights.
This new technology has raised questions about written description and enablement. Furthermore, AI systems raise issues relating to novelty, utility, and written description. Nevertheless, patents for AI systems are possible – and there are a variety of ways to protect an AI invention. In some cases, AI systems can be protected through a trade secret. A patent is also an important means of establishing the legitimacy of an invention in the market.
Inventorship of AI systems
The patentability of AI-generated inventions raises a number of interesting questions. Are they inventor-created or are the results of an AI process? As AI systems become more integrated with our daily lives, the debate on the issue will only intensify. Policy-makers should consider the underlying rationales for the patent system, including the promotion of innovation and the incentives to make full disclosure. Policymakers should also consider how AI can change these processes.
One recent case focuses on the question of who should be granted inventorship of AI systems. A court in Australia decided that AI systems should not be able to claim patent rights unless they are created by a human. The court’s reasoning is complicated, and it may take a few more cases before the issue is resolved. In the meantime, the AI industry must find a way to protect the intellectual property rights of AI systems.
Whether AI systems should be granted patent protection is an open question, but the legal system currently requires that humans be the inventor of all creations. This is problematic because a patent protects a human, but the AI system is a machine. Furthermore, if a human is not involved in the creation process, the patent office could refuse to grant IP rights for the creation. However, AI inventorship may be the right path for companies, despite the uncertainty surrounding it.
While many countries reject AI patents, South Africa and Australia have recently accepted AI as an inventor in its own right. The South African patent application will include the word ‘DABUS’ in the inventor’s name line. A case like this is a major breakthrough in AI innovation. The artificial intelligence behind these inventions has the potential to change the world. They will create more efficient systems for every aspect of life.
The legal status of AI-generated innovations is a hot topic for patent offices. A team of researchers from the University of Surrey, UK, filed two patent applications for AI-generated innovations. They named the AI system as the inventor of these inventions. However, the USPTO rejected the application, concluding that patent laws only grant rights to natural persons, and thus the Surrey team’s applications failed. Inventorship of AI systems may not be possible in the near future.
