AI copyright "milestone" agreement: Google and OpenAI negotiate media organizations to pay for training content

Author: Wang Mei

Image credit: Generated by Unbounded AI tools

AI companies are likely to need to pay for the copyright content used to train the ChatGPT model in the future.

The world’s largest technology companies are in talks with major media outlets over a landmark deal to train artificial intelligence using news content, the Financial Times reported on Friday. **

In recent months, OpenAI, Google, Microsoft and Adobe have met with news executives to discuss copyright issues surrounding AI products such as text chatbots and image generators, the report said, citing people familiar with the matter. Publishers including News Corp , Axel Springer , The New York Times , and The Guardian are all in discussions with at least one tech company.

The deals could involve paying media outlets subscription fees for their content to develop the chatbot technology that underpins the likes of OpenAI’s ChatGPT and Google’s Bard, these people added. These discussions are said to be still at an early stage.

“Your AI violates my copyright”

Since ChatGPT became popular all over the world, the issue of AI copyright has been a concern. The training of AI models requires a lot of data and content, but many of them are copyrighted. Whether AI companies should pay for the use of these content has become a focus of debate abroad recently.

The Wall Street Journal reported in March that News Corp., which owns the New York Post, Barron’s Weekly, and The Wall Street Journal, is preparing to seek compensation from AI technology manufacturers such as OpenAI, Microsoft, and Google through legal procedures.

The report quoted a person familiar with the media alliance organization as saying that recently some news executives have been studying the extent to which their content has been used to train AI tools such as ChatGPT and Bard, and are exploring how to obtain compensation through legal channels. This person said:

“We have valuable content that is the result of human hard work, but that content is constantly being used to generate income for other people and must be compensated.”

According to the analysis, the emergence of AI tools has exacerbated the already tense relationship between large technology companies and the publishing industry. Publishers have long relied on the help of tech companies such as Google and Meta to reach a wider audience with their content. But at the same time, a growing number of publishers are asking tech companies to pay for the use of their content.

** Publishers are concerned that AI tools could drain traffic and advertising dollars from their sites. **Media executives want to avoid the mistakes of the early Internet era, when many people made articles available online for free, ultimately undermining their business models. Big tech groups like Google and Facebook then used that information to help build multibillion-dollar online advertising businesses.

Not only the text copyright dealers, but the image copyright dealers have long been angry for not paying for the training of the ChatGPT model. **

In February this year, the American image trading company Getty Images sued Stability AI in Delaware, accusing it of infringing image copyrights; Class action. Earlier, in November last year, a federal court in California also filed a class action against Microsoft and its subsidiaries GitHub and OpenAI’s GitHub Copilot system.

In 2022, Dall-E 2 launched by OpenAI, Stable Diffusion created by Stability AI, and the AI image tool of the same name released by Midjourney set off a wave of image AI.

Among these AI tools, many works are generated by imitating the style of some photographic images or well-known paintings. In order for algorithms to emulate these styles, the companies that build them first have to copy the creations from the internet and then use them to train AI models. As a result, these companies have also been involved in one lawsuit after another due to copyright issues.