Amazon is in discussions with publishing executives about launching a marketplace that would allow publishers to sell their content directly to companies building artificial intelligence products, according to The Information.
The initiative would position Amazon as an intermediary in the intensifying dispute over how AI developers access, license and pay for digital content.
The proposed marketplace comes as publishers push back against the growing use of their articles, images and archives by generative AI systems. Media companies argue that AI chatbots and AI-powered search summaries are reducing traffic to their websites, weakening advertising revenue and undermining long-standing distribution models built around search and referrals.
Amazon Web Services has previewed the idea to publishers ahead of an AWS-hosted event in New York, per the report. Slides circulated ahead of the conference reference a content marketplace alongside core AWS AI offerings, including its Bedrock platform and productivity tools used by enterprise customers, according to The Information.
If launched, the marketplace would place Amazon in more direct competition with Microsoft, which rolled out its own AI content licensing marketplace last week. Microsoft said it has been testing the model by using licensed publisher content in both business and consumer versions of its artificial intelligence assistant Copilot before opening the platform more broadly to buyers. So far, Yahoo is the only publicly named content buyer on Microsoft’s marketplace.
Publishers increasingly favor usage-based compensation models that scale with how often AI systems rely on their content, rather than flat licensing fees. Industry executives say such models could offer a more sustainable revenue stream as artificial intelligence usage grows, but many also worry that AI companies may not participate in sufficient numbers to make marketplaces economically meaningful.
Amazon has already entered into direct licensing agreements with select publishers. The company is reportedly paying more than $20 million per year to The New York Times to use its content for training AI models and powering features in Alexa. Last week, Amazon also launched a free web-based version of its Alexa+ assistant, which incorporates content from more than 200 media outlets, including major newspapers and magazines.
Publishers are simultaneously turning to technical controls to limit unauthorized AI access. Infrastructure providers such as Cloudflare and Akamai have introduced tools that allow publishers to block AI crawlers or charge for access, while AWS offers similar capabilities through CloudFront. Even so, publishers say enforcement remains difficult, as some AI bots disguise their activity to resemble human traffic.
For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.