🔗 Share this article

An AI firm based in London has won in a landmark judicial proceeding that addressed the lawfulness of AI models using vast amounts of copyrighted material without authorization.

Judicial Decision on AI Training and Copyright

The AI company, whose leadership includes Oscar-winning director James Cameron, successfully defended against allegations from the photo agency that it had violated the global image agency's intellectual property rights.

Legal experts consider this decision as a setback to copyright owners' exclusive right to profit from their creative work, with one prominent attorney warning that it indicates "Britain's current copyright system is not sufficiently robust to protect its creators."

Findings and Trademark Concerns

Judicial evidence showed that Getty's images were indeed used to train the company's AI model, which allows individuals to create visual content through text instructions. However, the AI firm was also determined to have infringed Getty's brand marks in some cases.

The judge, Mrs Justice Joanna Smith, stated that determining where to strike the equilibrium between the interests of the artistic industries and the AI sector was "of very real societal importance."

Judicial Challenges and Dismissed Allegations

The photo agency had originally sued Stability AI for violation of its intellectual property, alleging the technology company was "completely indifferent to what they input into the development material" and had collected and copied countless of its photographs.

However, the company had to withdraw its initial IP claim as there was no proof that the training occurred within the UK. Instead, it continued with its suit claiming that Stability was still employing reproductions of its visual content within its systems, which it described the "core" of its operations.

Technical Complexity and Judicial Reasoning

Highlighting the intricacy of AI copyright cases, the agency fundamentally argued that the firm's image-generation model, called Stable Diffusion, amounted to an infringing copy because its development would have constituted copyright infringement had it been conducted in the UK.

The judge determined: "A machine learning system such as Stable Diffusion which fails to retain or reproduce any protected material (and has not done) is not an 'infringing reproduction'." The judge declined to rule on the passing off allegation and found in favor of some of the agency's arguments about brand infringement involving digital marks.

Sector Reactions and Future Consequences

Through a official comment, the photo agency said: "We continue to be deeply concerned that even financially capable companies such as Getty Images face significant challenges in safeguarding their artistic output given the lack of disclosure requirements. Our company committed substantial sums of pounds to reach this stage with only one company that we must proceed to address in a different forum."

"We encourage governments, including the United Kingdom, to establish more robust disclosure regulations, which are crucial to avoid costly legal battles and to allow creators to defend their rights."

The general counsel for the AI company commented: "We are satisfied with the court's decision on the outstanding allegations in this case. Getty's decision to willingly dismiss the majority of its IP cases at the end of court testimony left only a subset of claims before the judge, and this concluding decision ultimately addresses the IP issues that were the core issue. Our company is thankful for the time and effort the court has dedicated to settle the important questions in this proceeding."

Wider Industry and Regulatory Background

This ruling comes amid an ongoing discussion over how the current administration should legislate on the matter of copyright and AI, with artists and authors including several prominent individuals lobbying for enhanced safeguards. At the same time, tech firms are calling for wide availability to copyrighted content to allow them to develop the most powerful and efficient generative AI systems.

Authorities are presently seeking input on IP and AI and have stated: "Uncertainty over how our copyright system functions is impeding development for our artificial intelligence and creative sectors. That cannot persist."

Industry specialists following the issue indicate that authorities are examining whether to introduce a "text and data mining exception" into UK IP law, which would permit copyrighted material to be utilized to develop AI models in the UK unless the rights holder chooses their content out of such development.