Introduction
Welcome to the second quarterly Copyright and Generative AI (GAI) blog post of 2026. If you’ve come across this page, you’re likely working in the copyright field in some capacity. As such, the current debate surrounding what, if any, changes should be made to copyright law to account for advancements in GAI has likely piqued your interest. With so much information and so many opinions swirling around the internet from all sides of the debate, it can be hard to keep track of where the conversation stands. We hope these blogs will help make sense of it all.
Every three months, you can expect an updated post that includes important cases to watch, recent international developments, a summary of the latest discussions (with an effort to remain impartial) and finally, a suggested long-form article to help you dive deeper into one aspect of the subject.
So, without further ado, let’s get into it!
Cases to Watch
Cruz et al v. Anthropic, Case No. 3:26-cv-04482 (Northern District of California, 13 May 2026) USA
With the settlement in the Bartz v. Anthropic case now headed for final approval, we have started seeing several class action suits from individuals who opted out of the larger case. These authors, including those who joined Cruz's lawsuit, felt the larger (and much lauded) settlement to be unfair for many reasons. First, the amount of the settlement to be given to the teams of lawyers for the plaintiffs was argued to be disproportionately greater than what the actual individual injured parties would receive (roughly $3,000 per book infringed). There were also concerns about the scope of the settlement being too narrow, focusing only on the pirated works used to train the model and not the larger question of whether using copyright-protected works to train models could be considered fair use. Finally, choosing to remain in the larger class meant authors could not choose to sue Anthropic individually in the future for the same reason and potentially receive larger compensation packages and maybe even a clearer answer on the fair use question. With several of these individual cases now before the courts, it will be interesting to see how these new classes of authors fare against the largest of the GAI companies.
Thomson Reuters Enter. Centre GmbH et al v. ROSS Intelligence Inc., Case No. 1:20-cv-00613-SB (Court of Appeals for the Third Circuit, 24 June 2026) USA
We have also seen a familiar case resurface on appeal: Thomson Reuters v. ROSS Intelligence. You will likely recall that last year, the lower court found in favour of Reuters' argument that ROSS Intelligence infringed its copyright by scraping Westlaw headnotes in order to create a competing legal research product. ROSS has since appealed, and the case is now being deliberated in the Third Circuit, where the defendant is arguing that their particular use is transformative. The original case was hailed as a landmark decision regarding AI training (though, importantly, ROSS's product was not generative AI), so the copyright community is watching this appeal closely.
SURYAST U.S. ENTERPRISES, LLC v. Perlmutter, Case No. 2:26-cv-4999 (Central District of California, 8 May 2026) USA
In our last update, we spoke about Stephen Thaler's unsuccessful attempt to get the Supreme Court to hear his case against the US Copyright Office (USCO) for the applicability of copyright to his work, A Recent Entrance to Paradise. Another artist is also suing the USCO over its refusal to register his GAI work. However, this time, the artist in question is arguing that the image was merely enhanced by AI rather than being created autonomously by a GAI model as Stephen Thaler's work was. The work in question, called Suryast by its creator Ankit Sahni (an Indian lawyer), is a mashup of a photograph he took and Vincent Van Gogh's famous Starry Night painting. The Copyright Office refused registration of the work, claiming that it did not contain enough elements clearly created by a human; but Sahni is arguing that he not only took the underlying photo but also used GAI as a tool to make creative decisions. As with artist Jason M. Allen's case regarding his Théâtre d'Opéra Spatial work, it will be interesting to see if these cases lead to a ruling on the threshold for human creativity.
Yomiuri Shimbun v. Perplexity (Tokyo District Court, 7 August 2025) Japan
Though filed almost a year ago, this case has yet to be decided and is the first instance of a major media company in Japan suing a GAI company for unauthorized use of copyright-protected materials. This is not, however, the only reason for noting it here. It's equally important because it shows that it's not only North American media companies objecting to the use of their content in the training of GAI models. Similar to many of the North American cases, the plaintiff in this one is complaining that both its right of reproduction and the right to transmit to the public have been infringed by the services offered by Perplexity. Interestingly, in the Japanese context, there was an amendment made to the copyright law back in 2018 to accommodate machine learning and data analysis without requiring authorization. However, the amendment stipulated that this action could not "unreasonably prejudice" the rights holders of the content. This case, therefore, is the first in the country to test the limitations (or lack thereof) of this exception when it comes to GAI.
There are many other cases we could highlight here — 107 globally, with 81 of them based in the US, all of which will play a crucial role in determining the relationship between IP and GAI moving forward. You can see an excellent overview of all the cases currently before the courts globally here. It will be interesting to see if any of the cases mentioned above make it into our next roundup in one form or another.
International Developments
In our last quarterly update, we spoke about the Canadian government releasing the results of its public consultations about the future of AI in Canada. Since then, the federal government has held two summits, during which announcements were made regarding the future of AI in general in the country. In May, Minister of Artificial Intelligence and Digital Innovation Evan Solomon announced Can$66 million in funding for over 40 AI projects, as well as a new data centre to be built in British Columbia. Addressing the creators' perspective, Solomon hosted a summit in March, which disclosed the creation of an AI and culture advisory board to help ensure that any steps forward with GAI will keep creators in mind in terms of compensation and permissions. It's a step further than many nations have taken, including the US and the UK, and it will be fascinating to see where these discussions lead.
Some copyright commentators have pointed to the World Trade Organization (WTO) for possible answers to the convoluted relationship between copyright and GAI. This suggestion follows the European Parliament's adoption earlier this year of a resolution on copyright and GAI, which proposed that European law apply to every GAI model on the European market, regardless of where training might occur. The resolution predicts the defence most of the largest tech companies would mount, namely that their models are trained in the US; however, the resolution itself is non-binding. The resolution also called on the EU Intellectual Property Office to mandate a registry of the works used to train AI models, as well as a compulsory remuneration process. What the European Parliament seems hesitant to consider is that, legally, its work is domestic in nature and does not apply internationally. This is where the WTO's Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) may come into the picture. Within copyright circles, people are now wondering if Article 13 of the TRIPS Agreement, which lays out three specific conditions for the applicability of copyright exceptions, might provide an answer for the international community. With no clear solutions in sight, all possibilities must be considered!
We spoke above about how an Indian creator is currently suing the USCO for refusing to allow him to register his GAI work. While this regards the possibility of registering GAI works in the US, Sahni's home country of India is also grappling with whether or not works produced using GAI can be protected by copyright. Recently, Stephen Thaler, whose GAI-authored work A Recent Entrance to Paradise was denied copyright registration in the US, made a bid to register the same work under Indian copyright law, prompting the Delhi High Court to direct the Registrar of Copyright to consider his request. Similar to the case in the US, it is expected that Indian copyright law will be found to be incompatible with the concept of autonomously generated AI works. However, we will have to wait and see the final decision to confirm this suspicion.
Finally, perhaps copyright-adjacent but still noteworthy, a court in Germany recently found Google liable for false statements in its AI Overviews. The case in question involved two publishers who found that Google's AI Overview incorrectly associated their businesses with scams and other questionable business practices. They did not sue until after Google had failed to correct this error following its receipt of their cease and desist letter. Considering that this new element of Google's search function has been all but foisted on users internationally, whether they find it helpful or not, this ruling has serious implications for Google’s new interface. Since Google is certainly not the only company to have developed AI-powered search engines and chatbots, fear of liability might force multiple developers to shift their strategy moving forward.
Summary of Latest Discussions
Some of the latest discussions centre on existing exceptions in national copyright acts which are now being looked at from the perspective of GAI. This was mentioned briefly above in reference to Japan's machine learning exception, which may or may not apply to GAI's particular method of training. Similarly, the UK is taking another look at one element of its Copyright, Designs and Patents Act of 1988 which deals with authorship in "computer-generated" works. The interpretation when applied to GAI-created works is arguably complicated, leading some to argue that this particular provision is headed for repeal. If this were to happen, copyright in works produced using GAI would likely be decided on a case-by-case basis, leaving their fate overall uncertain. The article linked above goes through some interesting types of cases, including GAI-produced works based on underlying protected works, works created using complex prompts, and serial prompting’s potential to shift copyrightability. Considering the UK government seems to have adopted a "wait and see" approach overall, we must also wait to see whether the provision for computer-generated works will indeed be repealed or if it might be refined in the future to apply neatly to GAI-generated works instead. Similar existing provisions in South Korea and Europe (in the case of newer text and data mining exceptions) are being discussed as possible ways to regulate GAI moving forward.
Copyright-adjacent topics such as deepfakes and the use of AI to alter existing images without permission are controversial aspects of this new technology that continue to be discussed internationally. In the Netherlands, there was recently a situation in which a court illustrator's drawing of two Syrian brothers on trial for the murder of their sister was altered without permission by a Dutch far-right party to make the brothers appear more menacing. The party in question, Party for Freedom, ended up agreeing to pay damages to artist Petra Urban for allegedly infringing her moral rights. Elsewhere, Cyprus has amended its Copyright and Related Rights Law to introduce a new sui generis right that prohibits the imitation of a natural person's physical characteristics or a performer's performance. Basically, they seem to be taking the same tack as Denmark in trying to stop deepfakes through using copyright law as opposed to privacy or personality rights. It's clear that GAI has caused issues for not only the traditional economic rights covered by copyright law but related rights as well — so much so that we may need to add an extra session to this newsletter in the future! Do you think related rights deserve their own section? Let us know!
Deep Dive
Since publishers and authors from various fields account for so many of the companies and individuals currently scuffling with GAI, for this quarter's deep dive we're suggesting Jane Friedman's excellent AI and Publishing: FAQ for Writers. Well-researched and thorough, this post covers everything from how to copyright AI-assisted works to proving or certifying human authorship. While geared towards writers, this guide is helpful for anyone currently exploring AI's utility or even those still debating whether its usefulness is worth the inherent risks at this time. We hope you find it helpful!
More Information
That’s it for this quarter’s update. However, if you’re looking for more regular and up-to-date news about the nexus of copyright and AI, check out our free newsletter, Copyrighting AI. And if you have any thoughts on what we have shared here, please feel free to reach out!
