Forget “Google Zero.” We need to talk about “People Zero.”

This essay was originally published as part of NiemanLab's Predictions for 2026. Read Paul Cheung's 2026 predictions for 2026 here.

Your carefully curated homepage is invisible. Your design choices don’t matter. Your branding gets reduced to a small icon or a parenthetical citation in a chatbot’s answer — if you’re lucky to be cited at all. Pageview-based ad revenue collapses, along with your affiliate clicks. Email subscriptions stagnate with no new people seeing your sign-up forms. This is the world of “People Zero.”

The term “Google Zero” has been gaining attention to describe how publishers are losing traffic as people seek answers from AI tools like ChatGPT, Gemini and Claude, rather than sifting through links in search results and clicking to websites.

Follow that trend to its logical conclusion — combined with web browsers and operating systems increasingly building in AI features — and it won’t be long before few human beings show up to websites at all. This isn’t some five-years-from-now future. With AI chatbots regularly topping app store download charts and growing as the default way people explore the internet, this shift is well under way and will accelerate in the coming year, bringing us to “People Zero.”

So what’s a publisher to do? There are (at least) two paths, which can be pursued separately or together: First, leverage the new AI gatekeepers; and second, build digital products that go beyond just serving content.

First: Leverage AI gatekeepers

If you’re a major publisher, one of the big AI foundational model developers may come to you asking for a data partnership. The Associated Press, Condé Nast, The Atlantic, The Financial Times, Axel Springer and others have already done such deals. These agreements can generate some revenue and help ensure that AI systems reflect higher-quality journalism, but likely won’t be more than a small part of a broadly diversified revenue mix. For small publishers, your datasets aren’t sizable enough to justify major investment from an AI company or provide much meaningful revenue.

But there are other ways to work with AI platforms. One approach is offering an integration for your subscribers using Model Context Protocol (MCP), an open standard rapidly adopted across the AI industry. MCP lets AI systems talk to external data sources and applications — what some refer to as a universal USB-C cable for plugging into AI. Users would authenticate their accounts to connect the publisher’s MCP server to the AI chatbot client application.

A local news outlet could provide restaurant and event listings, campaign finance data or investigative archives. A niche trade publisher might offer industry databases or code compliance tools. When a user asks the AI for something relevant — “What’s happening with the new zoning proposal on Court Street?” or “Which health tech startups raised seed rounds this week?” — the app would know to pull from your data as the trusted source.

AI bots can also use “tools” provided through MCP, meaning publishers could provide services for users to take actions like booking a restaurant reservation, filing a building permit request, contacting a local representative or analyzing data. All of this offers new revenue possibilities for publishers.

Second: Build digital products — and experiences — that go beyond content

With the rise of AI, people increasingly expect their software and devices to adapt to their needs, remember their personal context and provide value they can’t get from a generic feed.

People don’t want broken logins, bloated ad tech and apps that forget who they are and what they care about. And most of us certainly won’t download a mobile app that’s just a wrapper for static content. Mobile apps themselves will become endangered as AI interfaces and agents take over tasks that can be done more efficiently by computers talking to other computers.

Publishers who thrive will offer experiences their audience simply can’t get from a generalized AI model. That includes:

• Connections to other people: Real communities built around a place or topic — whether through online groups, member networks or in-person events.
• Unique datasets: Tools, dashboards, trackers, or visualizations that help people make decisions about their job, health, money, education, local government or what to do in their free time.
• Immersive media experiences: Interactive or explorable stories with maps, simulations, timelines or narrative pathways that pull people into the material in ways AI summaries cannot.
• Direct communication channels: Personalized SMS, messaging, or push tools that cut through the noise — though these must be genuinely relevant or they, too, will be filtered by AI before a human ever sees them.

Journalism’s future is utility

“People Zero” pushes journalism to reconsider what we offer as our product. Paper technology dictated the static article as the end product, and publishers largely just ported that format to the web.

It doesn’t mean every news story must become a tool. Deep reporting and great writing remain valuable because these foundational journalistic practices inform, entertain, reveal injustice or show us the human beings behind the headlines. Readers will still seek out the voices of writers who they feel a connection with, whose work they admire and follow.

But alongside these stories, there’s a huge opportunity the industry has barely tapped: utility. The very AI technology that’s causing the disruption will also be how we can sustainably build these new personalized products. New AI software coding tools make it easier and cheaper to create digital experiences that previously were only feasible for the biggest tech companies.

AI can help journalism turn expertise into personalized guidance, and turn local reporting into action. It can transform information into services people rely on — things no generic AI model can replicate.

Burt Herman is principal and co-founder of Hacks/Hackers.