In the latest battle between AI and the media, major Danish newspapers and television stations are threatening to sue OpenAI unless the company compensates the country’s press for allegedly using their content to train their models.
“We want remuneration for our work (that) they have used to train their model,” says Karen Rønde, executive director of the Danish Press Publications Collective Management Organization (DPCMO), which represents 99 percent of the media. Danes, including state broadcaster DR. and TV 2. Rønde says the DPCMO plans to sue if an agreement is not reached within the next year.
AI has created a new front in copyright law after a series of demands It claimed that OpenAI, which is backed by Microsoft, crawled news companies’ websites without permission to train its AI models. Shortly after those lawsuits, OpenAI reached a series of licensing agreements with major publishers, allowing the company to train its future iterations of ChatGPT on their content. Financial terms of the agreements have not been disclosed.
Now, Danish media is trying to force OpenAI to negotiate with them as a collective, an unusual tactic that could provide a model for other small countries if successful. So far, OpenAI has closed deals with publishers individually and announced content partnerships with the Financial Times and the Atlantic, as well as German media conglomerate Axel Springer, French newspaper Le Monde and Spanish group Prisa.
After meeting with OpenAI online and in person earlier this year, Rønde says he was left with the impression that Denmark was not a top priority. “It became clear that the focus was the agreement in Germany, the agreement in France, the agreement in Spain and, of course, the American agreements,” he says. “There are so many content creators in every other territory and now they are left with nothing.”
Rønde sent a letter to OpenAI’s lawyer at Dutch firm Brinkhof informing them about Danish copyright law and says he is waiting for a response. He assumes that OpenAI has already used content from Danish press websites because the company has not told him otherwise, he says. Neither OpenAI nor Brinkhof responded to WIRED’s request for comment.
For Rønde, time is of the essence. He wants to strike a deal with OpenAI and also Google’s Gemini next year, before the use of AI chatbots and search engine summaries further marginalize publishers’ websites. “Maybe then it will be too late and the value of content for newspaper publishers will be, in one, two or three years, too low,” she says. “If we cannot conclude a partnership agreement within a reasonably short time, then we will have to enforce our rights.”
DPCMO was created in 2021 to help Danish media negotiate with Big Tech. “We needed to stay united, otherwise we were afraid that Denmark would be too small a country to have priority in the discussion with big tech,” says Rønde.
Last year, the group secured preliminary licensing deals with Bing and Microsoft’s Google to include Danish publishers’ content in the company’s search engines. Although the agreements stated that the two companies must compensate the publishers, they could not agree on how much.