One may be surprised to learn that in 2020, according to a study on intangible assets undertaken by Ocean Tomo (Intangible Asset Market Value Study – Ocean Tomo), 90% of the value of S&P 500–listed companies was accounted for by intangible assets, such as rights, intellectual property (IP), data, goodwill and software, as opposed to actual physical objects.
Creating and protecting this value depends in part upon laws that were conceived to confer a level of monopoly for a period of time over industrial and commercial processes and inventions (patents and designs)—plus the creation, performance, distribution and reproduction of original creative or access-controlled works (copyright and similar IP). These are areas in which JAMS neutrals assist in resolving cases through alternative dispute resolution (ADR) methods, including arbitration and mediation.
The Evolution of Laws
Some of these laws have adapted over time to try to catch up with technology. For example, the first copyright law in the U.S. was the Copyright Act of 1790. In England, the Statute of Anne was enacted in 1710 to protect authorship of books. The arrival of large-scale computing and data content led to the introduction of some extensions and bolt-on legislation. This legislation made it clear, for example, that authorship applied to creating gaming software. In the U.K., this was achieved in part by the Copyright, Designs and Patents Act 1988. However, when we talk about machines using generative AI to create their own work independently of human input, then all current laws are likely to be challenged.
Humans must produce the data that AI uses for its output. Without human input, the “well” from which AI draws would eventually dry up. Therefore, is the owner of the capital that led to something created by AI the “owner” of that creation and thus entitled to obtain protection?
Who Owns AI-Generated Works?
Interestingly, two countries, the U.K. and Ukraine, appear to offer some recognition of possible protection for AI technologies, but I suspect in the case of the U.K. that this is more an issue of interpretation of the Copyright, Designs and Patent Act 1988. This therefore places the U.K. in a position to decide whether the Act offers protection for computer-generated works. If so, will the protection be granted to the AI developer or the author whose work is facilitated by the AI model?
Courts in the U.K. and other countries must decide who specifically is recognized by most copyright law as the person “by whom the arrangements necessary for the creation of the work are undertaken.” The answer will turn on evidence. That evidence will require the unbundling of machine output from its original human input. Deciding then whether the human input is sufficiently material to meet the test set out in the existing domestic legislation will involve a decision. JAMS neutrals are likely well equipped to determine and resolve such issues efficiently and cost-effectively.