This Week in AI — 📰 What does AI hold for journalism in 2026?

What You Should Know

What Does AI Hold for Journalism in 2026?

Brands and news outlets see AI through different lenses. Brands are lining up to invite AI to scrape their websites in hopes of including them in the responses those systems generate. They’re already shifting strategies and optimizing content to match how AI search functions, because the stakes are growing as AI search becomes more popular. McKinsey projects that $750 billion in U.S. revenue will funnel through AI search by 2028, and companies want to be visible when those results appear. 

News outlets want to be compensated if an AI scrapes their site, because that process informs future models. Unlike traditional organic search, which returns a list of links, AI responses present a summary of what it learned in its search. It’s a zero-click world now. Even if an AI cites its sources with links, most users don’t actually click on them, which presents a problem for publishers. And publishers are struggling — each of the top 10 news websites reportedly saw year-over-year traffic decline in October, some exceeding 30%. In 2025, the mantra seemed to be “if you can’t sue them, join them.” 

The New York Times sued OpenAI and Microsoft in 2023, an ongoing case that alleges the companies were using millions of Times articles without permission or payment. Last week, it sued Perplexity in a similar case. But outlets are also willing to strike licensing deals, which the Times did with Amazon in May. Last week, Meta inked deals with several publishers, including USA Today, People Inc., CNN, Fox News, and others.

But the relationship between the industry and technology goes deeper than that. Journalism wants to use AI just as much as AI wants to use it. Newsrooms already use AI to automate content aggregation and scale research efforts. They’ll continue using AI tools in more ways as the industry’s next chapter plays out, which means PR teams will increasingly write for a hybrid audience. Those tools are becoming the first readers of many press releases. If your materials aren’t structured clearly or lack the context these systems need to categorize and resurface them, they’ll never reach a human audience. 

In fact, some think that’s the future. Daniel Trielli, Assistant Professor of Media and Democracy at the University of Maryland’s Philip Merrill College of Journalism, predicts that 2026 will bring the rise of “agentic journalism,” a new format that’s geared toward an audience of machines instead of humans. (Check out NiemanLab’s full collection of predictions for journalism in 2026 — it’s always an interesting read.) In Trielli’s view, stories may shed their classic shapes. Instead of narrative ledes or tightly woven arcs, pieces could break into bulleted lists or other formats that agents can ingest, summarize, and redeploy without losing meaning. 

If his prediction comes to pass, the ripple effects will hit communications teams. The clearer the scaffolding, the more your material stays intact when it’s scraped or reformatted into an agent’s output. Structure becomes the top priority. Provide clean facts. Anchor claims with sources. Offer context in ways a person can skim and a machine can index. Journalism might be inching toward a machine-first workflow, but if that’s the case, communicators won’t be helpless in getting their messages in front of the right audience.

Elsewhere …

• LISTEN: AI Cuts Payment Errors and Fraud by 70%
• Google to Launch First of Its AI Glasses in 2026
• ChatGPT Can Now Use Adobe Apps to Edit Your Photos and PDFs for Free
• OpenAI, Anthropic, and Block Join New Linux Foundation Effort to Standardize the AI Agent Era

Tips and Tricks

 Think deeper about instructions 

What’s happening: Tech companies and power users claim AI can do almost anything. They make it look easy because they show the results, but rarely highlight what’s under the hood: their prompts. In many ways, that gives the non-power users a false sense of what the technology is capable of.

Like what?: Take the stories about an AI-generated country hit that topped the charts last month. Some stories mention that it was created using a platform called Suno, but none get into how to actually use the platform or what prompts have led to the rise of AI music. A Wall Street Journal feature on the producer who built AI actress Tilly Norwood is a notable exception, but it only mentions the initial prompt that generated a result the producer didn’t like.

What to do about it: In the early days of AI chatbots, one popular prompting method was to convince the AI that it was an expert in what you were trying to get it to do. It wasn’t particularly effective. Saying “You’re the best writer that ever lived. Charles Dickens has a poster of you on his bedroom wall” might emulate Dickens’ style, but it won’t actually make the result good.

While we may not know the exact prompts used for the most popular AI creations, three-plus years of using the technology tells us that the more detailed the instructions, the better the result. Instead of saying “Write a newspaper column on Topic X” or “Edit this piece,” try using more descriptive language, like “Provide feedback on this piece against the following criteria: whether it follows AP style, whether it’s repeating words or points, and whether it would align with the interests of X audience.” 

Keep notes on the prompts that work well and those that don’t in a library. Here’s a head start for some common tasks communicators have, but you’ll want to make a copy and tailor it to your own needs.

Quote of the Week

“In 2026, LLMs will focus journalists on the work people want us to do — and on building relationships with those who want to have them with us.

“There are more ways than ever to understand and respond to what people value. And we have learned that serving our audience is the best path to success on the business side.

“As AI continues to evolve, it’s going to be more important than ever that we listen. Next year, I’m not going to worry about the doom people think AI will rain down on journalism. I am going to focus relentlessly on the jobs people want and need us to do for them.”

— Lauren Gustus, CEO and Executive Editor of The Salt Lake Tribune, in her NiemanLab prediction for journalism in 2026

How Was This Newsletter?