AI and Media liability – Keep calm and Carry On

By Angela Weaver, Focus Group Leader - Media & Entertainment at Beazley

In legal binds, where art and craft doth dwell,

The scribes and artists to their works lay claim,

With copyright's shield, their rights to affirm and tell,

That none may steal, nor tarnish their good name.

The above is the beginning of a Shakespearean style sonnet about copyright law, generated by an AI chatbot in seconds. Shakespeare’s works are in the public domain so no issue there, also the Bard did not write anything, as far as we are aware, specifically on the topic of copyright law. But, the capability of the AI chatbot imitating Shakespear’s style is clear. 

Artificial Intelligence (AI) and Large Language Models have now evolved to a point where it is difficult to distinguish their output from works created by humans. An author’s style and output can now be impersonated with increasing accuracy in seconds. This is not limited to the written word. Music, photography, and art can all be replicated in the style of a particular author, musician, or artist, with increasing likeness. The tools to do this are now everywhere and are becoming more sophisticated. 

The media industry is, understandably, concerned. Our Risk & Resilience research shows that a fifth (20%) of media, telecoms and technology businesses cite failing to keep pace with new technologies as the biggest threat to their businesses this year, with 28% anticipating this to be the case in 12 months’ time.

This boom in AI has begun to test the bounds of copyright law and prompted introspection across the creative sector. Are the days of authors, musicians and artists numbered? Will their ability to protect their works from plagiarism be removed? Thankfully, the short answer is no. Despite AI's transformative impact, copyright law's core principles remain relevant.

To understand why, we need to remember that we have been here before. AI is not the first technological evolution to test copyright law or insurance   against infringement. Rather, last year, marked a quarter century since the beginning of the music industry’s long war with the internet. 

The evolution of copyright infringement

In 1999, Shawn Fanning created Napster, a peer-to-peer file sharing application which allowed users to share music over the internet. The copying of music, often to cassettes or CDs, which also constituted copyright infringement, was not new. What made Napster different was the scale of it. The internet was now enabling copyright infringement on an industrial scale. The music industry at the time was in uproar. If music could be shared and obtained online for free then no one would pay for it. Metallica's drummer, Lars Ulrich, filed a lawsuit in 2000, alleging copyright infringement.¹ Other artists and record labels swiftly followed suit. 

Napster was, ultimately, accused by numerous plaintiffs of copyright infringement and eventually found liable. The site was forced to close in 2002, apologised publicly, and paid millions in damages. This process has since been repeated multiple times over the past two decades. The advent of new technology leads to litigation, which leads to legislation and regulation, or the development of new precedent to protect creators. 

When YouTube launched in 2005, it rapidly became a hub for video sharing, attracting millions of users worldwide. However, this burgeoning popularity soon caught the attention of copyright holders, leading to significant legal challenges that would shape the future of online content sharing. Organisations from Mediaset to the English Premier League filed lawsuits against YouTube, claiming that it had done little to prevent the uploading of copyrighted material.²

The most prominent legal battle in YouTube's early history was the lawsuit filed by Viacom in March 2007.³ Viacom, a media conglomerate owning numerous television channels and film production companies, accused YouTube of copyright infringement. After years of litigation, the  U.S. District Court for the Southern District of New York granted summary judgment in favour of YouTube, holding that the company’s video hosting operations fell within the “safe harbour” of the Digital Millennium Copyright Act (DMCA).⁴ This landmark ruling had profound implications for the sharing of content online. 

Identifying copyright infringement in AI-generated works is complex. AI models can produce vast content, obscuring specific infringements and complicating the determination of intent. This may make AI copyright infringement actions more difficult and costly. However, the legal principles that apply to traditional content creation still govern AI-generated content. The rapid advancement of technology invariably outpaces the evolution of legal frameworks. However, this does not mean that businesses can use AI generated content with impunity or that it is now open season on artists IP.