Amid the increasingly intense copyright lawsuits and data hunger in the artificial intelligence industry, Amazon plans to use its vast cloud services (AWS) ecosystem to build a "legal trading bridge" between publishers and AI developers.
According to a report by The Information on Monday, Amazon has begun promoting a new content trading market to publishing executives. In a slide shared internally ahead of the AWS Publishers Conference on Tuesday, the platform's concept was revealed: publishers can directly list their content assets (such as articles and archives) on the market and license them to technology companies developing AI products.
From "Passive Crawling" to "Transparent Market"
For a long time, the acquisition of AI training data has been in a gray area. While companies like OpenAI have avoided legal risks by signing individual agreements with organizations such as the Associated Press and News Corp, this "one-on-one" negotiation model is difficult to scale.
Amazon's Model: It plans to integrate the market with its AI tools such as Bedrock (Basic Model Service), allowing developers to access compliant and high-quality training materials directly in the cloud.
Pioneer in the Industry: Microsoft has also recently launched a similar "Publisher Content Marketplace" (PCM), aiming to provide a transparent economic framework that allows publishers to define licensing terms themselves.
A Lifeline or a Temporary Fix for Publishers?
Media organizations are facing an unprecedented traffic crisis. Recent research shows that "AI summaries" provided by search engines like Google have caused a sharp drop in website click-through rates (CTR), with some sites losing up to 25% to 40% of their traffic.
New Business Model: Publishers tend to view this market-based system as a more "sustainable" model compared to simple licensing agreements.
Past Cases: Amazon has already shown ambition, reportedly paying over $20 million annually to The New York Times for AI training and summary display in products like Alexa.
