Google is asking news publishers to grant broad content usage rights, including for AI training, as a condition of joining a new pilot program for AI-powered article overviews in Google News and the Gemini chatbot. Publishers who refuse will lose the annual fee they currently receive for appearing in Google News, a program Google plans to end, The Information reported Thursday (June 25, 2026), citing an unnamed source.
Those rights could allow Google to use participating publishers' content to train its artificial intelligence models. A Google spokesperson said "as people's news preference change, we've been expanding our partnerships through our News AI pilot program, working with a wide range of publishers to explore how AI can drive more engaged audiences." The company said in a December blog post that it was updating its partnerships with more than 3,000 publications "for the AI era."
The pilot program and what Google wants
The pilot offers publishers new features such as AI-generated article overviews on their Google News pages, designed to give readers more context before they click through. In exchange, Google is seeking broad content usage rights that could extend to training its AI models. The exact scope of those rights has not been publicly detailed, but the arrangement signals a shift in how platform companies value and compensate news content.
Publishers face a stark choice: participate and grant the requested rights, or see their existing revenue stream from Google News disappear. The current program that pays an annual fee for article display is set to end, and the new pilot is the only path forward for those payments.
Regulatory and legal pressures mount
The European Commission announced in December an antitrust investigation into whether Google used web publishers' content to power generative AI features in search results without proper compensation or an opt-out mechanism. That probe adds to the scrutiny facing tech companies that train AI models on publicly available data.
Separately, a coalition of nearly 400 local and regional newspapers filed a lawsuit Wednesday against OpenAI and Microsoft, alleging copyright infringement. The suit claims the companies used the newspapers' copyrighted articles to build and train commercial AI products without permission or payment. These cases underscore the growing legal friction over what constitutes fair use of published content in AI training pipelines.
Why this matters for IT, development, and government professionals
For IT and development teams, the story highlights a critical data provenance issue: understanding where AI training data comes from and whether it was obtained with proper consent can affect model compliance and risk. Organizations building or deploying generative AI tools-such as those covered in Generative AI and LLM courses-must now track not just performance but the legal standing of their training datasets.
Government and policy professionals are watching the antitrust investigation and copyright lawsuits as early indicators of how regulators will treat AI training data. The outcome could shape rules for public-sector AI procurement and for how agencies use commercial models. Professionals who work with Google's ecosystem may also want to review Google AI Courses to stay current on the company's evolving tools and partnership terms.
Your membership also unlocks:
