Yahoo boss warns: AI is a 'threat to publishers' very existence' as copyright battle heats up

"AI companies rely on vast quantities of data, including books, images and journalism, much of which is protected by copyright. Media groups argue this is an unauthorised takeover of their work that threatens their revenue. Several lawsuits are under way, including a high-profile case by The New York Times against OpenAI. Some firms, including Reddit and Rupert Murdoch's NewsCorp, have signed licensing deals with AI developers."

"Jim Lanzone, who runs the American online publishing giant, said Yahoo was one of the most heavily "pilfered" sources of material used to train AI models. He criticised the way AI systems scrape articles without permission, only to repackage them for users without linking back to the source. "Unlike search, where the business model was an understood agreement - the engine aggregates and then sends traffic downstream to the publisher - the AI model takes content without consent. It's like signing away your future existence,""

"Yahoo, which remains one of the top five most-visited websites in the world, takes stories from agencies like Reuters and AP alongside its own journalism. Its model relies heavily on ad revenue driven by traffic - which Lanzone said is being eroded by AI "short-circuiting" the link between readers and publishers. Since being taken private by Apollo Global Management in a $5bn deal in 2021, Yahoo has tried to strengthen its role as a major aggregator and content provider."

Yahoo identifies its content as heavily pilfered for training AI models and criticises AI scraping that repackages articles without linking back to sources. The practice contrasts with search engines that historically sent traffic downstream to publishers, instead taking content without consent and undermining future revenue. Media groups characterise large-scale training on copyrighted books, images and journalism as an unauthorised takeover that threatens income, and several lawsuits — including a high-profile case by The New York Times against OpenAI — are under way. Some companies have pursued licensing deals, but publishers view these largely as damage control. The loss of referral traffic is eroding ad-driven business models.

Read at Business Matters