Mat Kenyon
Spotlight 08 Informed
At Risk: Our Creative Future, led to the scheme being dropped before the general election. Now, amid vigorous Silicon Valley
lobbying that is arguing the UK risks being leſt behind in the evolution of AI, the idea appears to be back on the table. A government consultation on the issue is expected to be announced before Christmas. It will seek to resolve the growing
tensions between data-hungry AI developers and content creators, including journalists, who believe the tech sector is taking without paying. “It is a bit like asking a builder to build your house but refusing to pay for the materials,” says Owen Meredith, chief executive of the News Media Association, which represents national newspaper publishers. Following rampant scraping of online
Big Tech’s Content Grab
Ian Burrell reports on what a text and data mining exception could mean for journalists.
Across the creative sector, there is growing concern that the government plans to bring back a proposal that could allow tech giants including Alphabet and Microsoſt to exploit the work of UK content owners for training their AI models without paying. It was thought these controversial plans to expand the text and data mining (TDM) exception to UK copyright law were dead in the water. Te Conservative administration, when envisioning post- Brexit Britain as a light-regulation AI hub, embraced the idea of ripping up the rules to allow AI developers free access to copyrighted content and databases for “any purpose”, including commercial ones. But a fierce backlash from the media and arts sectors, in addition to a critical report by the House of Lords Communications and Digital Commitee,
copyrighted content, both before and aſter the revolutionary launch of OpenAI’s ChatGPT model in 2022, several publishers have elected to sign licensing deals for the use of their material. Te Financial Times and Reuters are among those working with Microsoſt’s Copilot model. Conde Nast is partnering with OpenAI, which has also signed a five-year deal with NewsCorp – reportedly worth $250m over five years – and is offering upwards of $1m-a-year to smaller news organisations. Te existence of these deals is cited
by Dan Guthrie, director general of the Alliance for Intellectual Property, as proof that “there is no market failure” in UK copyright law and therefore no need for an extension of the TDM exception. But such deals are happening only in “very limited circumstances”, he says, because tech firms generally want to train their large language models (LLMs) for nothing. “Tey need huge access to data and they claim it would be too complicated to ask permission so they are very keen to scrape lots of content from across the web without having to seek consent and therefore licensing and, therefore, paying for it.”
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12