Artificial Intelligence and the Law: Who owns the assets created by an AI robot and what does it mean for Canadian business?
Many successful start-ups find success utilizing new technology to solve old problems, or to make things easier and more convenient. One such technology is artificial intelligence (AI). As AI becomes more common, it is important to understand the legal risks and ambiguity that exist. A recent Interesting Engineering article raises an important question; who owns intellectual property (IP) created by AI?
The World Intellectual Property Organization, located in Geneva Switzerland, has an online exhibit featuring the art of Ai-Da, an AI humanoid robot who has sold its works for over $1.5 million in Canadian dollars and also appeared in a music video for the band The 1975. The AI program used by Ai-Da utilizes sight via camera vision, neural networks, and a paintbrush or pencil in her hand to complete its art. The AI program itself was developed by PHD students from Oxford University. The exhibit focuses on the ways AI can transform culture and industries, showing that robots like Ai-Da are likely to continue creating works or products which will make the question of who owns their work more relevant.
The exhibit explores the question of how, “In a world that is increasingly governed by algorithms, many of which function completely under the radar, where does human governance end and robotic self-ownership begin?” As technology continues to advance, this will only become a larger question. Within this general question the exhibit wonders if AI can be a creator or inventor within current intellectual property frameworks, or if a human is required. The exhibit even questions what AI is since no agreed upon definition seems to exist. This is further complicated by the fact even human intelligence can be difficult to clearly define.
While Ai-Da may be one of the more unique and independent AI artists, it is far from alone. Google, of course, has both a program which would help to write local news articles, as well as software that generates unique music by listening to recordings. The news articles still involve significant human capital and inputs, and to a lesser extent the music program also is reacting to what humans are training it with, but they show first steps towards how AI could become very prevalent in society and the economy. Interestingly, a novel co-authored by an AI program almost won a Japanese literary prize competing against human authored novels. In this case, the human team selected sentences and words while setting parameters for the program. It then completed the novel titled, “The Day A Computer Writes A Novel” when translated to English.
With new developments, many have begun to argue that intelligent robots need to be treated more like human beings. Some believe that robots in the workplace deserve ethical rights as their use, and violence against them, increases. A more detailed, and legal, case is made by Ryan Abbott in “The Reasonable Robot: Artificial Intelligence and the Law.” Abbott argues that the law should not discriminate between AI and human behaviour.
In Canada, it is far from clear who would own IP created by an AI program. The Copyright Act requires “the author was, at the date of the making of the work, a citizen or subject of, or a person ordinarily resident in, a treaty country.” This would not cover AI, or if it were classified as a computer, potentially leaving their works exploitable by others. However, Canadian law has not directly confronted the emergence of AI as a potential author and creator.
There are many routes that can be taken for regulating AI and how IP and employment laws apply to it. One potential avenue that would eliminate risk for entrepreneurs is to treat robots and AI like property, specifically how the law treats pet ownership. If something like Ai-Da earns money or other positive benefits, then the owners of Ai-Da could reap the rewards. Alternatively, if Ai-Da somehow injured people or caused property damage, those same owners could face the consequences. This is certainly not a perfect fix, but it is just one example of how the issue could be handled without drastically changing Canadian law. Artificial intelligence is well on its way toward being a major feature of everyday life and regulations and laws need to adapt and be proactive instead of reactive for when, inevitable, something goes wrong with AI and ends up in the courts as a dispute.
Other countries have already made determinations on how they will handle AI creations. The United States copyright office has followed U.S. court decisions in determining that an original work of authorship must be created by a human being. Australian courts similarly determined work generated with the intervention of a computer could not be protected by copyright because it was not human produced. Alternatively, jurisdictions like New Zealand, Ireland, India, Hong Kong, and the United Kingdom have gone with a different approach. The best example is that of the U.K., which through legislation has declared that “In the case of a literary, dramatic, musical or artistic work which is computer-generated, the author shall be taken to be the person by whom the arrangements necessary for the creation of the work are undertaken.” They go further in defining computer generated work, clarifying it “is generated by computer in circumstances such that there is no human author of the work”, which creates an exemption to any human requirements.
Artificial Intelligence is not going away. It is going to continue to evolve and create more headaches for those responsible for policies and laws that govern its use. Regardless of the route Canadian lawmakers take, let’s hope businesses and entrepreneurs can receive some certainty about how IP laws will treat AI creations.
 Chris Young. (2020, September). New Exhibition Explores Whether AI Robots Should Own Intellectual Property. Retrieved from https://interestingengineering.com/new-exhibition-explores-whether-ai-robots-should-own-intellectual-property.
 WIPO and IP: A virtual experience. Retrieved from https://wipo360.azurewebsites.net/.
 The 1975. (2020, July). The 1975 - Ai-Da responds to ‘Yeah I Know’. Retrieved from https://www.youtube.com/watch?v=dTK9N7n8Wcg&ab_channel=The1975VEVO.
 Julia Gregory. (2017, July). Press Association wins Google grant to run news service written by computers. Retrieved from https://www.theguardian.com/technology/2017/jul/06/press-association-wins-google-grant-to-run-news-service-written-by-computers.
 Devin Coldewey. (2016, September). Google’s WaveNet uses neural nets to generate eerily convincing speech and music. Retrieved from https://techcrunch.com/2016/09/09/googles-wavenet-uses-neural-nets-to-generate-eerily-convincing-speech-and-music/.
 Chloe Olewitz. (2016, March). A Japanese A.I. program just wrote a short novel, and it almost won a literary prize. Retrieved from https://www.digitaltrends.com/cool-tech/japanese-ai-writes-novel-passes-first-round-nationanl-literary-prize/.
 Copyright Act, RSC 1985, c C-42. Retrieved from https://www.canlii.org/en/ca/laws/stat/rsc-1985-c-c-42/latest/rsc-1985-c-c-42.html.
 Andres Guadamuz. (2018, October). Artificial intelligence and copyright. Retrieved from https://www.wipo.int/wipo_magazine/en/2017/05/article_0003.html.
Blog posts are by students at the Business Venture Clinic. Student bios appear under each post.