Oct 5, 2026 · 39 min
AI training puts publishers’ fragile business model on trial
AI Companies Scraped the Web and Are Destroying the Human Internet
The dispute over scraped content reaches beyond copyright, threatening the audience, data, and advertising systems that finance digital journalism.
- 1AI companies’ alleged use of publisher work raises a new copyright fight over training, summaries, and competing products.
- 2Google’s AI answers could deepen publishers’ traffic losses after years of dependence on advertising intermediaries and audience data.
- 3Jason Kint argues publishers need paid, direct audience relationships alongside stronger legal protections for content and data.
Don't miss
Kint argues that publishers should be able to permit ordinary search crawling while refusing AI training and summaries, with opt-in licensing as the eventual remedy.
The brief
Newly unsealed records involving Microsoft and OpenAI frame AI training as an alleged appropriation of journalists’ work, turning familiar web scraping into a dispute over labor and value.
Jason Kint, CEO of Digital Content Next, traces the problem to publishers’ dependence on Google, Facebook, and advertising intermediaries that traded reach for data and revenue.
The conversation’s legal center is The New York Times’ case against OpenAI, where fair use, licensing, and the difference between search crawling and AI training remain unsettled.
Google’s AI Overviews add a second pressure point: publishers may lose traffic when search answers questions directly, even as their reporting helps power those answers.
Kint links the renewed Cambridge Analytica debate to broader accountability for platform privacy failures, then argues publishers must build direct, often paid audience relationships.
Featuring
The New York Times
OpenAI
Microsoft CorporationListen to the full episode and explore every guest, topic, and moment on PodLume.