ISOJ 2026 Day 2: Publishers are facing a new middleman as AI changes the business

The news business is confronting an existential paradox: AI technology can make journalism faster and cheaper but it also threatens the audience and revenue that sustain it.

That tension dominated the second day of the 27th International Symposium on Journalism, where news executives and journalists on Friday debated AI-powered workflows, new products, falling web traffic and the looming rise of AI agents.

At Axios, AI is helping local reporters produce regional news roundups in minutes rather than hours. The company is also developing AI-driven products, including a “things to do” guide for local audiences.

Politico reporters, meanwhile, are using AI to tackle massive datasets and even write code they don’t know how to write themselves — with help from chatbots and newsroom engineers.

But the harder question is what happens when AI becomes the destination rather than the tool. Traffic to news websites has plunged since consumer AI tools took off after 2022. And if AI agents eventually gather and summarize news for consumers, publishers face a new dilemma: How do you make journalism for readers who may never see it?

“We’re on the verge of a gigantic transformation of the news media ecosystem,” said Rosental Calmon Alves, founder and director of the Knight Center for Journalism in the Americas, which organizes ISOJ, during a panel he moderated on agentic AI.

Here are some of the future-looking ideas from the second day of ISOJ:

Reporters are using AI to power their investigations

Many top news organizations are working to build the news business of the future. That means embracing AI without letting it consume the industry it promises to transform, and serving younger audiences increasingly more via creators rather than traditional news organizations.

At Politico,two reporters used AI earlier this year to sort through more than 20,900 court cases involving ICE detention and turn the findings into a format accessible to readers.

The reporters manually compiled and categorized the records involving ICE detentions, then used AI to extract basic case information. Their analysis found that most of the rulings involved immigrants eligible for deportation who were detained without basic due-process protections.

However, the enthusiasm for the yet-untapped potential of AI comes with caution. Politico editor-in-chief Jonathan Greenberger, speaking at the “Journalism’s AI moment” panel, said he is also wary of what the technology could mean for the news business. The organization has blocked AI chatbots from scraping its content, he said.

All things considered, Greenberger said, “there’s never been a better time to be a journalist, not just because of the incredible story in front of us, but also because AI gives us incredible tools to cover this changing, disruptive world.”

New technologies are also changing how people consume news, creating a new challenge for legacy publishers to reach audiences.

Among people ages 18 to 24, that increasingly means social media. A Reuters Institute analysis published in March found that 51% of young people pay more attention to individual news creators rather than traditional news brands.

Adriana Lacy, chief executive of the Boston-based Field Nine Group, uses her business to connect social media creators with legacy newsrooms to help expand their audience reach.

“When we’re looking at creators for newsrooms to work with, we’re looking at who’s the transit advocate in your community that only talks about trains and bikes, and you have a story about how you want to do that,” Lacy said at a panel called “Reaching young audiences.”

Lacy added that the worlds of social media creators and newsrooms are increasingly the same. She once met a journalist from a large news organization who described herself as a “secret influencer” but did not want to be identified because she is not creating news content, she said.

Ryan Kellett, co-founder of the Independent Journalism Atlas, which tracks models to produce reporting trusted by consumers, described how newsrooms should recognize the potential for newsroom creators and attract new audiences.

“If you get to the point where you have an influencer in your newsroom, that is an amazing win for you,” Kellett said.

News companies are making money with AI

As AI continues to reshape how people find and consume news, executives from Axios and the Sembra Media news incubator discussed how news organizations can use AI to build new revenue streams and create a more sustainable business model.

Allison Murphy, chief operating officer of Axios, said the company has incorporated AI into its newsroom and, as a result, strengthened its hyperlocal reporting. Axios — a digital news publication that operates newsletters in 45 communities — is known for its “smart brevity” and usually has two reporters per community.

“The economics have to work because the core problem in local journalism is an economic one,” Murphy said. “How do you make the math work to be able to bring great information to communities?”

Murphy said AI tools are used to automate some tasks in the newsroom. For example, a news roundup section of their daily newsletter that used to take reporters three hours can be completed in about 10 minutes with AI tools, allowing reporters to spend more time on original reporting.

“We’re taking anything that isn’t a reporter who’s talked to a source, who’s dug into a story, anything else we want to make as automated as possible,” Warner said.

Murphy said the company is also experimenting with AI-generated products that could create new advertising opportunities. Axios is testing a “things to do” newsletter that is about 90% AI-generated. She said that during beta testing, audiences told the company they did not mind the use of AI as long as the content was good.

Murphy said, however, human expertise and connection are still at the forefront of their mission. One of the most popular parts of each newsletter is the reporters’ sign-off at the end, where reporters share personal details about their lives with readers.

“Readers constantly cite that is the thing they love the most: the human connection,” Murphy said. “That is something that the AI can’t replicate.”

For Janine Warner, executive director and co-founder of SembraMedia — an organization working to strengthen independent digital media across Latin America — the economic question is also central to how news organizations should approach AI. She urges news organizations to begin thinking about how they can monetize it.

“If we don’t start thinking about how to make money with AI now, we’re going to be left behind,” Warner said.

Warner highlighted various AI companies and tools that can help newsrooms with monetization. But she also warned newsrooms to be cautious about the risks of generative AI, like intellectual theft and deepfakes.

“I still believe that newsletters, events, things that really help us have direct connection, are also very popular right now because people want to connect with people,” Warner said. “If you’re going to be using AI, my best advice is let the good bots in, keep the bad ones out.”

What AI traffic and agentic AI will mean for news consumption

The rise of consumer AI tools such as ChatGPT, Claude and Gemini has sent traffic to news websites off a cliff, as Google puts AI-generated summaries ahead of classic search results and users increasingly turn to chatbots for information.

Panelists from outlets including The New York Times, ProPublica and the BBC discussed Friday how news organizations should manage AI-driven traffic and how agentic AI could reshape news consumption.

The decline in traffic has sent many publishers asking whether their teams are creating a product that only machines will read. And yet, they are also seeing a growth in traffic coming from readers arriving via “dark social” — in other words, stories that users share through email, private messaging, and text messaging.

“This is what the human audience is doing,” said Jill Nicholson, chief marketing officer at Chartbeat, which tracks website data and analytics.

Direct traffic is driven by breaking news content and fact-based, in-depth stories that could not be easily replicated by chatbots, according to Dan Petty, former director of audience strategy at ProPublica.

“Habits are changing, and changing quickly, and just indicate the importance of brand and the importance of building trust and being very clear about what your value proposition is to your audience”, Petty said.

The next frontier seems to be AI agents — systems that can search and act on information with little human prompting — but there is disagreement on whether it should be seen as an opportunity or a threat.

Florent Daudens, chief executive of the AI-native distribution platform Mizal, said some organizations are shifting from a business-to-consumer to a business-to-AI-agent model, where the information will be intermediated by an AI agent.

“If you treat the agent as a customer, then you start thinking about building a conversion funnel for the agent, the same way we build a conversion funnel for the humans,” Daudens said.

However, news organizations seem to be still skeptical about the impacts of Agentic AI. For Zach Seward, editorial director of AI initiatives at The New York Times, consuming news through a personal AI agent remains niche. But he said news organizations urgently need to adapt to other forms of AI, including chatbots.

“We need to pretty dramatically change how we publish information in the first place, connected to citations of supporting facts and allow our readers and agents to do their own research along the way,” Seward said. “If there’s a mass exodus from news sites to chatbots and AI summaries, journalism is going to get absolutely flattened by products cosplaying as objective omniscient oracles unless we get started working on a different way of doing things.”