My Lords, given the hour, I will be brief. That was an absolute tour de force by the noble Baroness. As with all the Minister’s speeches, I will read her speech over Easter.
I was very interested to be reminded of the history of Napster, because that was when many of us realised that we were, in many ways, entering the digital age in the creative industries and beyond. The amendments that the noble Baroness put forward are examples of where the Bill could make a positive impact, unlike the impact that so much of the rest of it is making in watering down rights. She described cogently how large language models are ingesting or scraping data from the internet, social media and journalism, how very close to the ingestion of copyright material this whole agenda is and how it is being done by anonymous bots in particular. It fits very well with the debate in which the Minister was involved last Friday on the Private Member’s Bill of the noble Lord, Lord Holmes, who inserted a clause requiring transparency on the ingestion or scraping of data and copyright material by large language models. It is very interesting.
The opportunity in the data area is currently much greater than it is in the intellectual property area. At least we have the ICO, which is a regulator, unlike the IPO, which is not really a regulator with teeth. I am very
interested in the fact that the ICO is conducting a consultation on generative AI and data protection, which it launched in January. Conterminously with this Bill, perhaps the ICO might come to some conclusions that we can use. That would of course include the whole area of biometrics, which, in the light of things such as deepfakes and so on, is increasingly an issue of great concern. The watchword is “transparency”: we must impose a duty on the generative AI models about the use of the material that they use to train their models and then use in operation. I fully support Amendments 103 and 104 in the name of the noble Baroness, even though, as she describes them, they are a small step.