People Inc. is building custom APIs to make its journalism easier for AI agents to use 

“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,” Roberts told delegates at Digiday’s Media Strategies London event. “It will really distract the agent, and it will be incredibly expensive to process all these tokens that it doesn’t need.”

The custom APIs strip out that extra material, sending the article text in Markdown along with details such as who wrote it and when it was published or updated and other “trust signals that matter.” 

The move comes as People Inc. shifts its broader investment priorities: the company laid off about 6% of its workforce last week. The cuts underscore the pressure on publishers to find durable businesses as AI reshapes how audiences discover and access journalism.

People Inc. first gave partners access to its content through web crawling, then through Amazon S3 buckets, which let the company send updates and retract content. It is now moving toward custom APIs, Roberts said, because crawling is an inefficient way for AI systems to get information. 

Strategically, that could make access more efficient and give People Inc. greater control over how its journalism is delivered to AI partners.

“Every token that needs to be processed is expensive,” said Roberts. “If you throw it the entire HTML header, that’s abuse. Right, that’s not reasonable.”

AI companies are beginning to recognize the importance of knowing when publishers created or updated content, Roberts said. People Inc. is building those details into its content for its own AI tools, then using what it learns to offer AI partners a more useful way to access that material. The publisher is not focused on advertising directly to AI agents, he added. 

The urgency comes as traffic from AI agents and agentic browsers grew 7,851% year over year, according to HUMAN’s 2026 State of AI Traffic report. 

The challenge is already pressing for publishers trying to distinguish useful, customer-led agents from bots scraping content. “We need to act quickly on it because it’s not going to be something in the future — it’s happening now,” Daily Mail director of SEO and editorial commerce Carly Steven said at Digiday’s Media Strategies London event. “So that is a fairly urgent need that we all need to address and get our heads around,” she added.  

Publishers are testing different ways to adapt their content and services for AI systems. People Inc. is moving toward structured access for AI partners. The Economist is experimenting with making marketing and B2B sales material outside of its paywall more easily readable for agents. Time has also converted its pages from HTML to Markdown and has even started to sell ads to those agents visiting its site. 

Alongside figuring out which automated requests to allow, publishers are experimenting with how to package their journalism for AI systems. Forbes is also testing how to make some content more usable by AI agents. For its Vetted product reviews, the publisher has created Markdown versions of its content and an “MCP layer” while exploring separate presentations for booths and human readers.

Nina Gould, Forbes’ chief innovation officer, said the approach could help the publisher reach people using AI tools, but emphasized that the strategy is still developing a planned ChatGPT product-recommendation plugin that has not yet gone live. “On something like ads for agents, we’re still working on that,” she added.