AI firms have attracted billions in investment and all big tech firms have pivoted towards the technology. Image Credit: Shutterstock

Paris: Artificial intelligence firms must ask for permission before using copyrighted text and images to generate content, a consortium of news groups said on Wednesday.

Media organisations including AFP, Getty Images and the Associated Press said in an open letter that AI tools risk shattering their business models, flooding the web with misinformation and breaking copyright laws.

AI tools like chatbot ChatGPT and image generators DALL-E 2, Stable Diffusion and Midjourney exploded in popularity last year with their ability to generate a wealth of content from just brief text prompts.

However, the firms behind the tools, including OpenAI and Stability AI, already face lawsuits from artists, authors and others claiming their work has been ripped off.

Wednesday's open letter from news organisations including European Pressphoto Agency and Gannett/USA TODAY is the latest attempt to influence the debate by organisations that have much to lose if AI firms continue to scrape material from the internet without restrictions.

Watermarking pledge

"Generative AI and large language models make it possible for any actor, without regard to their intent, to produce and distribute synthetic content at a scale that far exceeds our past experience," the news groups wrote.

They listed potential problems including copyright breaches, a flood of false or biased content, and the creation of a vicious cycle where media groups can no longer fund journalism to provide reliable information.

The organisations said they wanted to be part of the solution and called for talks to ensure legal access to content.

AI firms have attracted billions in investment and all big tech firms have pivoted towards the technology, with Google and Microsoft leading the way.

Major companies in the field have formed several coalitions - including the Partnership on AI and the Frontier Model Forum - and largely push for self-regulation.

They committed in July to taking measures like watermarking AI content and have broadly pledged to fight misinformation, but with no specific timetables or measures.

However, little has been said about copyright, which is likely to be a much thornier and potentially expensive issue.

Google used a public consultation in Australia earlier this year to call for a "fair dealing" exception in copyright law specifically to allow data mining for AI.