Innovation Partner Peter Nunn spoke with The Times on news that US comedian and author Sarah Silverman is suing the ChatGPT developer OpenAI for copyright infringement.
Silverman has claimed that OpenAI was trained on her memoir, The Bedwetter, without her permission. The lawsuit, which is also being brought by two other authors, Christopher Golden and Richard Kadrey, argues that OpenAI’s breach of copyright likely comes from websites that illegally aggregate content.
Peter Nunn spoke with The Times to discuss what impact any ruling could have for the training of artificial intelligence models going forwards. Peter commented: “If OpenAI is held to have infringed copyright in the books in the US, then it is likely that it may adjust generally how it uses people’s copyright works to train ChatGPT”.
Peter also noted the UK Government's current proposals for a code of practice and suggested that in future there could be a compulsory licensing model on rights holders, whereby they are paid a fee in return for the use of their work, in a similar model to the music industry.
Read in full (subscription required).