Google has been warned that reusing news articles to respond to queries from internet users may breach European Union copyright rules.
The EU Court of Justice Advocate General Maciej Szpunar said that chatbots that retrieve and reuse information from online publications to generate answers could fall foul of copyright laws.
“When an LLM (large language model)-driven chatbot retrieves a publication from a press publisher’s website, reuses that publication to generate an answer, and provides that answer to the user, the press publisher’s right may be infringed in certain circumstances,” Szpunar said.
He added that copyright rules do not protect “mere facts” reported by the media but a “specific form of expression”.
Szpunar’s statement is in response to a court case between Hungarian publisher Like Company and Google over its Gemini chatbot.
In the lawsuit, Like Company argues that Google used copyrighted articles to train its large language models and reproduced them through Gemini between June 2023 and May 2024.
Hungarian judges turned to the ECJ for guidance about the matter in relation to EU law.
A Google spokesperson told the Wall Street Journal: “We acknowledge the Advocate General’s opinion recognising that EU law permits AI training on publicly available information. Gemini learns linguistic patterns to create original outputs rather than replicating content.
In response, Google said that publishers have the ability to opt out of its AI features.
Publishers fear that opting out of Google’s AI features could penalise them on Google search, but the tech giant has previously stated that it would not impact their search ranking.
In June, the UK’s Competition and Markets Authority (CMA) mandated that publishers should have the right to opt out of Google’s AI features.
The issue highlights tension between publishers and tech companies that train their LLMs on news content.
Publishers argue that since AI-powered answer machines have grown in usage, website traffic has fallen off a cliff.
Many media organisations are working with LLMs on content deals that allow the platforms to use their content for a fee.
Since Google launched its News AI pilot program last year, more than 400 publications are taking part, including The Washington Post, Axios, CNN, The Financial Times and Forbes.
