Putting a price on the value of IP

The government sees artificial intelligence (AI) as a high octane fuel to boost productivity and drive the economy forward. The Regulatory Innovation Office (RIO) aims to cut bureaucracy to fast-track the rollout of new technology, which can often get hampered by red tape.

Labour’s proposal that paves the way for AI firms to use copyrighted material for training their models, was dealt a blow last week following a House of Lords proposal to the Data Bill, forcing AI companies to reveal which sources they used to train their AI models. The House of Lords supported by 287 votes to 118, an amendment to the Data (Use and Access) Bill, adding a commitment to introduce transparency requirements, aiming to ensure copyright holders are able to see when their work has been used and by who.

When asked by the BBC’s Laura Kuenssberg about the risks of allowing AI firms to train their models on copyright-protected works, Elton John said it would amount to “committing theft on the highest scale.” The singer-songwriter warned that AI could run rampant without artists having to ask for approval. He pointed out that artists generally don’t have the resources to keep checking for copyright infringements or the means to fight a lawsuit.

Responding to the Lords’ amendment, techUK has warned that it would create serious barriers to AI innovation and research and undermine the UK’s ambition to be globally competitive in AI development, negatively impacting economic growth.

The government wants to give the UK every chance of becoming an AI superpower, by providing a legislative framework to encourage innovation. But at what cost?

In spite of being global communities where anyone can showcase their work or art, the web and social media are dominated by a handful of extremely powerful companies that are 100% focused on the bottom line – which is to deliver shareholder value, usually by monetising their worldwide reach.

Earlier this year, Deepseek caused a furore, not only by being a Chinese-built LLM that more-or-less matched and in some cases, surpassed the best the US could deliver, but OpenAI claimed it stole the outputs from OpenAI models as part of the distillation used to improve its performance.

Tech firms always seem to cry foul play when someone takes their intellectual property (IP). Yet at the same time, the likes of OpenAI have grown rich by scrapping every bit of content that we freely post on the web and social media sites, to enrich their own data troves and line their pockets. The software industry is no different to the creative industries.

We all need to accept the importance of protecting IP, irrespective of what tech chiefs want us to believe.