The Emerging Media Lab at UBC hosted an EMCop session that delved into the complex and evolving intersection of intellectual property law and artificial intelligence 🤖. The discussion, titled “IP Law and Artificial Intelligence,” explored key questions surrounding ownership, copyright, and responsibility in the age of generative AI.
The session kicked off by examining fundamental IP concepts such as copyright 📝, patents 💡, trademarks ©️, and trade secrets 🤫, setting the stage for a deeper dive into how these apply to AI-generated content. A central theme revolved around the question of whether the tool or the creator holds the rights to AI-generated output, a query that remains largely unanswered in current legal frameworks.
Attendees learned about the historical context of copyright, including the Statute of Anne from 1709, and how copyright isn’t about outright ownership but rather a license to exploit creative work. The discussion also highlighted the significant impact of “Mickey Mouse” on copyright law, noting how the initial 50-year copyright term was extended to 70 years, largely due to efforts to protect iconic characters. The Bern Convention was also mentioned in this context.
A critical point of contention was the ownership of generative AI output. The session emphasized that current legal interpretations lean towards the human, not the AI tool, as the author. This led to a fascinating discussion about how much “poking” of the AI is required to consider the output human-authored. The evolving nature of copyright law was underscored, with examples like ChatGPT’s ambiguous stance on using YouTube videos.
The session also touched on the responsible use of AI, particularly in academic settings. It was highlighted that if AI is used for a paper, its use must be cited. Furthermore, the responsibility for any AI-generated mistakes falls squarely on the individual instructor, as per the Conference Board of Canada.
The conversation then shifted to contemporary challenges and emerging legal issues. The ongoing legal battle between ChatGPT and the New York Times 📰 was cited as a prime example of the complexities arising from AI’s use of copyrighted material. The troubling rise of AI-generated revenge porn 💔 and the passing of new laws to combat it were also brought to attention. It was noted that one cannot copyright a style.
The session concluded by emphasizing the importance of documenting the entire creative process to keep the human in focus, a point attributed to “Patrick”. The discussion provided valuable insights into the dynamic landscape of IP law as it grapples with the rapid advancements in artificial intelligence 🧠.