by Jeremy Erwin , Ross Magee April-26-2024 in Commercial & Business, Intellectual Property
Copyright infringement actions against AI providers in relation to training of generative AI are now gaining traction in the US. There are a number of high profile cases being brought against AI providers such as OpenAI for alleged copyright infringements.
The alleged infringements are based on the training of AI systems. AI systems are fed from a body of material on the internet including books, music, newspaper articles etc. to teach them. The AI providers are defending their use of these materials and therefore the proceedings “fair use” from Title 17 US Code Section 107. A fair use of copyright material is where it is used for purposes such as criticism, comment, news reporting, teaching, scholarship, or research.
A case going to trial to watch out for is Thomson Reuters v Ross Intelligence. Thomson Reuters, who own the legal research platform, Westlaw, accused Ross Intelligence, a legal research start up, of copying Westlaw headnotes and Key Number System to train its AI. Ross Intelligence argue fair use and that it “only studied the language patterns in the headnotes to learn how to produce judicial opinion quotes”. Thomson Reuters argue that “Ross used the untransformed text of headnotes to get its AI to replicate and reproduce the creative drafting done by Westlaw’s attorney-editors.” This case could set an early precedent for fair use in the field of generative AI.
Another case of interest involves the Authors Guild, who have joined a number of US authors including George R R Martin, in a class action suit against OpenAI for the use of author’s stories, characters and voices in training ChatGPT. This in turn allowed users of the AI system to create sequels such as the long awaited final instalment to the A Song of Ice and Fire series. OpenAI have added Microsoft to the complaint as they argue the training could not have happened without Microsoft’s financial and technical support.
Universal Music and other music producers are suing Anthropic for the misuse of “innumerable” of song lyrics to train its chatbot Claude. The music producers argue the chatbot is programmed to generate copyrighted lyrics.
Both ChatGPT and Claude were allegedly trained on inputs asking for reproduction of copyrighted materials.
The New York Times are also suing OpenAI for the use of millions of copyrighted news articles, investigations, opinion pieces etc. Interestingly, the New York Times note in their complaint that following their objection to the use of their materials, they sought to negotiate term for their use. This highlights a common point raised that the issue the copyright owners have is the infringement of their rights. They would not necessarily object to the use of their materials for training AI systems if they were involved in the negotiation the terms of their use.
An interesting point has been coming up recently in the IP world in relation to AI; who owns the IP rights where the AI system creates the material/inventions? In the US, a recent case has reinforced their position in copyright law that there must be human authorship. In Ireland, copyright law already provides that the copyright for work which is computer-generated is owned by “the person by whom the arrangements necessary for the creation of the work are undertaken.”
The US Patent and Trademark Office issued guidance on for ‘AI-Assisted Inventions’. This advice includes that inventors and joint inventors named on US patents and patent applications must be natural persons. Applicants should not list AI systems as joint inventors and a natural person who creates an invention using an AI system must “contribute significantly” to every claim set forth in the invention. There is no requirement to disclose to the USPTO that the inventor used AI as part of the invention process. However, this is only guidance from the USPTO.
A recent UK case affirmed that a machine which embodies an AI system could not be an inventor under the Patents Act.
The patentability in this field is untested in Ireland, however, there are a number of elements we can assume will be considered. Firstly, computer programs are not considered to be an invention under Section 9(2) Patents Act. Secondly, ‘Inventor’ is defined as “the actual deviser of an invention”, which appears to leave the question open. However, Section 80 of the Patents Act, refers to co-owners as “two or more persons”, in relation to co-ownership of patents. Lastly, the European Patent Office has stated that an inventor must be a person with legal capacity.
If patent protection is not an option for AI generated content, trade secrets is a possible alternative. The European Union (Protection of Trade Secrets) Regulations 2018 sets the criteria for an algorithm to be classified as a trade secret, there are three essential criteria:
- it must be actually secret;
- it must have actual or potential commercial value; and
- there must be reasonable efforts made to keep it a secret.
This quite clearly is a very important developing body of case law for authors, developers, musicians, and anyone else relying on intellectual property rights to protect their brand or identity. Given the impact AI is already having on day to day society, it is important for regulators and legislatures to be proactive. The above represents only a handful of cases that are progressing in various jurisdictions, so this is an important areas to watch.
Back to Full NewsShare this article:
About the Authors
Jeremy Erwin
Jeremy specialises in insolvency, commercial litigation and dispute resolution, acting for a variety of companies and financial institutions in contract law cases, enforcement and recovery actions and in high value complex Commercial Court proceedings. Jeremy also specialises in intellectual property matters, including advising on registration and protection of trade marks and related rights and on trade mark disputes.
Ross Magee
Ross is a Solicitor in the Commercial & Business team at Hayes solicitors. He practices predominantly in insolvency, commercial litigation and dispute resolution. He acts for a variety of corporates, financial institutions, private equity funds and private individuals in contentious matters that include, insolvency; enforcement and recovery; landlord and tenant; intellectual property, defamation and other general litigation claims.