Ad/Marketing Communication
This legal article/report forms part of my ongoing legal commentary on the use of artificial intelligence within the justice system. It supports my work in teaching, lecturing, and writing about AI and the law and is published to promote my practice. Not legal advice. Not Direct/Public Access. All instructions via clerks at Doughty Street Chambers. This legal article concerns AI Law.
Introduction
I don’t write about artificial intelligence and intellectual property as often as I’d like. It’s not my primary area of practice, but my interest was piqued following discussions with several individuals deeply concerned about where artificial intelligence and intellectual property law are heading, how they might protect their intellectual property, and what the future holds for creators and businesses alike.
My attention was particularly drawn some time back now when Jack Dorsey tweeted ‘delete all IP law,’ sparking immediate debate. Individuals, including Elon Musk, joined the conversation, raising serious questions about whether such radical proposals were more than just provocative statements. Could AI effectively erase an entire branch of law and if so, what would fill the vacuum left behind? Legal experts quickly responded, emphasising that abolishing IP law entirely would undermine incentives and severely destabilise creators’ rights.
Clearly, this debate is far from settled, and this past week (20–26 June 2025) has brought significant international developments. Several high-profile cases globally have illustrated how courts are cautiously navigating the complex and evolving relationship between artificial intelligence and intellectual property. While AI developers achieved partial victories in key jurisdictions, critical questions remain unanswered, setting the stage for further legal drama ahead.
BBC v. Perplexity AI (UK – 20 June)
In the UK, one significant development this week was the BBC’s move against Perplexity AI. On 20 June, the BBC wrote a letter accusing the AI startup of scraping and using BBC news content without permission to train its chatbot-like search engine. This move highlights the fundamental tension surrounding IP rights and AI training methods. The BBC, like many other media outlets, is concerned about protecting its content from unauthorised use and exploitation. Perplexity, however, disputes the claim, arguing that the BBC misunderstands the nature of generative AI technology.
Tucker v. Apple Inc. (US – 20 June)
Meanwhile, in the US, Apple faced a notable class-action lawsuit brought by investors alleging securities fraud. Shareholders accused the company of overstating the capabilities of its AI-powered Siri assistant, negatively impacting sales of the iPhone 16 and causing Apple’s stock to plummet by nearly $900 billion since late 2024. I always find it difficult to navigate where AI claims are highly exaggerated or where perhaps they have been understated. This case is a timely reminder that AI exaggeration isn’t just harmless enthusiasm it carries real financial and legal consequences. Those of us that are approached with AI legal tools may wish to follow this closely. The case can be read here.
Bartz et al. v. Anthropic (US – 23 June)
Perhaps the most significant legal development this week came from a landmark ruling in California involving Anthropic. Judge William Alsup determined that the company’s use of copyrighted books to train its AI model was ‘transformative’ and therefore fell under fair use. However, this victory was tempered by the court’s criticism of Anthropic’s methods of obtaining training data from illegal “shadow libraries.” A further trial set for December will address these remaining claims. This nuanced ruling illustrates that while AI training can be lawful under fair use, the methods of obtaining training data remain closely scrutinised. The case can be read here.
Bird et al. v. Microsoft Corp. (US – 24 June)
Only one day later, authors filed a substantial copyright lawsuit against Microsoft in New York. The case alleges that Microsoft used a vast dataset of pirated e-books to train its AI model, ‘Megatron,’ effectively copying the authors’ writing styles without permission. The timing of this lawsuit, coming immediately after the Anthropic ruling, may prove significant. There remain important unresolved questions and uncertainties around fair use principles, and we could still see different courts arriving at diverging conclusions regarding AI’s relationship with intellectual property. The case can be read here.
Authors v. Meta Platforms (US – 25 June)
Another key decision from California arrived as Judge Vince Chhabria dismissed a similar authors’ copyright claim against Meta, ruling that the plaintiffs had not provided sufficient evidence of market harm. Although this was seen as a partial victory for AI developers, the judge clearly indicated this ruling did not set a broad precedent, explicitly noting that future plaintiffs might succeed by presenting stronger evidence. Again, a conditional legal success. The case can be read here.
Getty Images v. Stability AI (UK – ongoing)
Here in the UK, the Getty Images drama continues, and I’ll be writing a full post about this, including some of the “fun” pleading issues, at a later date. It was one of the first cases I addressed on this blog which can be read here.
However, what the claim is about, in brief, is Stability AI’s use of millions of Getty’s photographs in training its AI image generator, Stable Diffusion, without a licence or permission. The key question the court must tackle is whether existing IP law is flexible enough to handle situations where AI models learn from copyrighted materials but produce new and different works. This case is important because it could establish clearer boundaries around how traditional copyright and trademark laws apply when AI is involved, potentially setting significant precedents for many similar disputes to come.
Legislative and Regulatory Developments
I know the main focus of this post is legal cases, but it’s worth briefly highlighting some legislative and regulatory developments internationally as well.
- United States: There’s a controversial proposal to block local AI regulations for ten years, but it’s currently facing strong opposition from both parties, clearly, there’s no consensus on the best way to regulate AI
- India: Regulators have issued proposals calling for more transparency and stronger oversight of AI used in financial markets, reflecting a wider global trend towards cautious regulation.
- United Kingdom: Auditing bodies here have published new guidance encouraging auditors to use AI responsibly and transparently.
Overall, these developments underline just how uncertain and complex the regulatory landscape around AI remains, with governments everywhere still grappling to find the right balance between innovation and oversight.
Conclusion
I am still very concerned about the “delete all IP law” discussions and what they represent. Those who know me, know I class myself as a musician, albeit not a great one (my kids are far better), but I genuinely enjoy creating original music. It’s not easy, and it takes significant time, so it would be a serious concern if that content could not be properly protected.
I’m particularly troubled by the idea of abandoning protections that allow creators to maintain control over their content and ensure fair recognition for the considerable effort involved. IP laws exist to protect creativity, encourage innovation, and acknowledge that producing original work genuinely requires time and skill. Removing those protections entirely feels reckless, especially when I think about independent artists who rely heavily on these safeguards.
On the other hand, I’m realistic and do appreciate that artificial intelligence and intellectual property must coexist and evolve. AI has already trained extensively on existing creative material, undeniably challenging traditional boundaries and forcing us to confront difficult questions about what creativity actually means, and whether our current legal frameworks remain fit for purpose. I’ve experimented with creating music using AI tools myself, and while I found the results impressive, it also made me acutely aware of how quickly and easily original creative works can now be generated.
As this week’s developments clearly demonstrate, AI does seem to be gaining ground in these IP battles, but its victories remain cautious, conditional, and far from certain. I’ll be following closely as artificial intelligence and intellectual property continue their uneasy dance, sometimes complementing each other, sometimes colliding, while we try to work out where human creativity fits in a world where machines can compose, paint, and write at the click of a button. We desperately need a legal framework that respects both the transformative potential of AI and the irreplaceable value of human originality. Without that balance, we risk eroding the very foundations that make art, music, and expression worth protecting in the first place.
What are your thoughts on this issue? The discussion continues on LinkedIn and my Substack and don’t forget to subscribe to my newsletter here for more insights and updates on emerging AI legal challenges.
