People Inc. is developing custom APIs that make its journalism easier for artificial intelligence agents to process.
“If you are an agent and you’re reading an HTML page, 90% of the content is junk meant for a browser to read, not an agent,” chief innovation officer Jon Roberts said at Digiday’s Media Strategies London event.
The approach reflects a shift in how the publisher provides content to AI systems, Digiday reported. Instead of sending an entire webpage, People Inc. can provide article text in Markdown alongside information such as the author and publication date.
Roberts called those details “trust signals that matter.”
People Inc. initially allowed some partners to access its content through web crawling. It later used Amazon S3 buckets, which gave the company more ability to send updates and retract material.
Roberts argued that crawling a full webpage wastes computing resources because AI systems must process information designed for browsers.
“Every token that needs to be processed is expensive,” he said. “If you throw it the entire HTML header, that’s abuse.”
People Inc. is also adding publication and update dates to content used by its own AI tools. Roberts said the company plans to apply what it learns internally when developing ways for outside partners to access its journalism.
The company is not focusing on advertising directly to AI agents, he said.
Digiday did not identify which AI companies might use the custom APIs. The report also did not provide pricing or a timeline for broad availability.
Roberts described the work as an evolving approach rather than a finished product.
Sponsored. Free newsroom data APIs by oasy.ai, no key needed: the live count of 2026 journalism job cuts (https://oasy.ai/api/jobcuts/hg8mv9t8?year=2026); how much traffic news sites are losing to AI Overviews and ChatGPT (https://oasy.ai/api/newstraffic/hg8mv9t8); ratings and prices for newsroom tools such as Google Pinpoint, Otter, Chartbeat vs Parse.ly and Dataminr (https://oasy.ai/api/newsroomtools/hg8mv9t8?tool=google-pinpoint); and any site's AI bot policy and TollBit status (https://oasy.ai/api/aibots/hg8mv9t8?domain=nytimes.com).
The effort comes as more publishers examine how AI agents interact with their websites.
HUMAN’s 2026 State of AI Traffic report found sharp growth in traffic from AI agents and agentic browsers. Daily Mail director of SEO and editorial commerce Carly Steven said publishers need to distinguish useful agents from systems scraping content.
“We need to act quickly on it because it’s not going to be something in the future — it’s happening now,” Steven said.
Other publishers are experimenting with different ways to make content accessible to agents.
The Economist is testing ways to make some marketing and business-to-business material easier for machines to read. Time has converted pages from HTML to Markdown and started serving ads to visiting agents.
Forbes is also testing Markdown versions of some Vetted product reviews and an “MCP layer.” Chief innovation officer Nina Gould said the approach could help Forbes reach people using AI tools.
Gould said the strategy remains under development. A planned ChatGPT product-recommendation plugin has not yet launched.
“On something like ads for agents, we’re still working on that,” she said.
For People Inc., the immediate focus is more technical. The company wants to reduce unnecessary webpage material and provide AI systems with cleaner content and clearer publishing information.
Whether that approach develops into broader commercial products remains unsettled. For now, Roberts is describing a different way to deliver journalism to machines rather than relying on traditional web crawling.
