Artificial Intelligence – Questions of ownership

Current IP laws are not well suited to deal with the situation where autonomous AI creates potential intangible assets, because in many cases those assets can only arise if there is a human creator. Patent law generally considers the inventor as the first owner of the invention. The inventor is the person who creates the invention. In the case of autonomous AI generating an invention, there is no legal owner as the AI technology cannot own the invention.

Intellectual property laws and systems establish rules about who owns which intangible rights. These rules have generally been developed and introduced to facilitate commerce and trade as parties can buy, sell and license rights in a way that is generally clear to them. There are also well-established ways of resolving disputes about the intangible rights. Another argument supporting intangible intellectual property rights is that they stimulate research and innovation. The cost of research is high and investors will not pay that cost without a reasonable chance of a return on their investment. The laws and systems enabling these intangible rights to exist are sometimes complex and expensive. However, the prospect of not having intellectual property rights is one of chaos and uncertainty, which is bad for business and the economy.

The current IP laws and systems do not offer an answer to a situation where IP rights cannot protect assets that are a product of autonomous AI. It is also not sensible or practical to continue with an approach where no one owns the potential intangible assets created. The situation is generally the same in many countries around the world. For instance, US law states that “Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title,” which implies that a person needs to have made the invention, not a computer.

Thinking about how we can adapt intellectual property laws to better cope with assets created by autonomous AI, one option is to enable the autonomous AI itself to own the intangible assets. Those who argue that giving autonomous AI the status of a legal person would address the issue of accountability,
effectively imply this solution. The question of accountability is identifying who is responsible for things that go wrong because of the use of autonomously acting artificial intelligence technology.

If a fully autonomous robot, such as Rachel in Bladerunner, has the status of a legal person, then it follows that she can be an inventor and subsequent owner of a patent. In the same way, she could be the author of copyright in a computer program or the creator of a reputation, which gives rise to passing off rights for unregistered trade mark protection. Therefore, it seems impossible to answer the question of ownership of intangible assets created by fully autonomous AI without at the same time answering the question of accountability. The two issues are inextricably linked. However, it seems inappropriate to give autonomous AI the status of a legal person. Humans should have the ethical duty to take responsibility for any autonomous AI technology they commission, deploy or use.

Alternatively, future law could give autonomous robots a status akin to that of a child. A human would then be responsible for the robot in the same way a human parent or guardian is responsible for a child. This goes some way towards addressing the problem of accountability. A significant limitation of this
approach is that the robot would own the intangible assets whereas the human parent or guardian would have very limited control.