The artificial intelligence company behind ChatGPT has launched its video generation tool in the UK amid a deepening row between the tech sector and creative industries over copyright.
Beeban Kidron, the film director and cross-bench peer, said the introduction of OpenAI’s Sora in the UK added “another layer of urgency to the copyright debate”, in a week the government faced strong criticism over its plans for letting AI firms use artists’ work without permission.
San Francisco-based OpenAI is making Sora available to UK users who pay for ChatGPT. The tool stunned film-makers when it was revealed last year, with the film and TV mogul Tyler Perry pausing an $800m (£634m) expansion of his Atlanta studio complex after saying the tool might make building sets or travelling to locations unnecessary. It was launched in the US publicly in December.
Users are able to make videos on Sora by typing in simple prompts such as asking for a shot of people walking through “beautiful, snowy Tokyo city” where “gorgeous sakura petals are flying through the wind along with snowflakes”.
OpenAI announced the UK release as it released examples of Sora’s use by artists from across the UK and mainland Europe, where the tool is also being released on Friday. Josephine Miller, a 25-year-old British digital artist, created a two-minute video of models wearing bioluminescent fauna and said the tool would “open a lot more doors for younger creatives”.
However, Kidron said the launch underlined the importance of the debate over copyright and AI in the UK, which centres on government proposals to let AI firms use copyrighted work to train their models – unless creative professionals opt out of the process.
“Comments made by YouTube last year make clear that if copyrighted material was taken without licence to help train Sora it would have breached their terms of service. Sora would not exist without its training data which means it is built on stolen goods. At some point YouTube may want to take action on that,” she said.
Last year the head of the video platform said it would be a violation of its terms of service if YouTube content had been used to train Sora. However, asked if YouTube clips had been used in this way, Neal Mohan told Bloomberg: “I don’t know.” The chief executive added: “It does not allow for things like transcripts or video bits to be downloaded, and that is a clear violation of our terms of service.”
The Guardian reported on Tuesday that ministers were considering offering concessions over copyright to certain creative sectors.
after newsletter promotion
Sora also offers users the option to make clips of varying lengths, from five to 20 seconds, which can then be extended to make longer videos. Features include displaying the clip in a variety of aesthetic styles, including “film noir” and “balloon world” where objects are represented as inflatables.
Clips can take a minute to generate at a low resolution and four minutes or longer at a higher resolution. A “storyboard” option allows users to tweak the video by editing a more detailed version of the prompt created by the underlying AI model that powers Sora.
OpenAI said use of copyrighted material to build Sora complied with copyright law and the tool was built using a wide range of datasets, including publicly available data. The company, which also announced the latest version of ChatGPT on Thursday, admitted last year that it would be impossible to create tools like its groundbreaking chatbot without access to copyrighted material.
