As AI pulls users away from search engines, how do consumers find their way to news sites? If AI agents increasingly replace human visitors to Web sites, how can media companies monetise these new consumers?

During a Product & Tech Seminar at the 2025 World Congress of News Media, leaders from prominent AI companies discussed how the technology is disrupting the traditional relationship between the media industry and consumers — and what that relationship may look like going into the future.  

Balancing accuracy, attribution, and AI partnerships

Jodie Hopperton, lead of the INMA Product & Tech Initiative, introduced representatives from major AI players who held a panel discussing their approaches to working with media companies.  

Nikhil Kolar, vice president, AI, for Microsoft in the United States, argued that AI providers need to forge partnerships with news outlets to acquire data inputs and ensure the accuracy of outputs by AI models. 

“I don’t think we are trying to be the arbiters of truth,” Kolar said. “We rely on partners.” 

Jessica Chan, head of publisher partnerships at Perplexity in the United States, spoke to the importance of AI platforms properly attributing the sources from which their answers come.

It’s not just about respecting the content rights of media organisations; AI services cannot function without the information funnelled into their models from publishers. This mutual dependence means that both the media and AI industries have vested interests in supporting each other.

“In order for us to be successful, we need publishers to be successful,” Chan said. “We’re not here to take jobs, we want to make your job easier.”

Surviving the AI revolution

Brian Alvey, chief technology officer at WordPress VIP and Gabriel Koen, senior vice president, tech at PMC, discussed how media brands can thrive amid AI-fueled upheaval. 

Alvey explained that, while many media industry leaders are keenly aware of the transformative potential of AI, they feel that the money they’re investing in it isn’t yielding sufficient results. 

“Media companies are saying they invest just enough in AI to stay behind,” Alvey said. “They’re not winning, they’re not leading, they’re not keeping up. They’re investing like crazy, but they don’t feel like they’re keeping up.”

For Koen, the best way to take advantage of AI is to use it to augment successful operations and tactics that media outlets have already been pursuing, rather than trying to shoehorn in novel features.  

“A lot of where we have seen the most success with using AI is to supercharge things that are already happening,” Koen said. “For the newsroom, that’s tools that are helping them get their work done faster. Things like surfacing better related content.”

Where media companies encounter disappointment, Koen cautioned, is when they attempt to integrate novel AI features into their products that don’t enhance content output or the user experience: 

“What’s really interesting is a lot of the experimentation we have done around what might be a killer AI feature doesn't really have a lot of traction within our newsrooms,” Koen said.

“I don’t think it’s necessarily even a reluctance on anyone’s part to change how we’re doing things. I think it’s that we’re doing a lot of things, as product and tech or as writers and journalists, that are working and effective.”

One problem, two solutions

In the final session of the seminar, two vice presidents from different companies described their approaches to ensuring AI does not compromise on obtaining appropriate permissions, attribution, and monetisation for content owners. 

Josh Stone, vice president, business development at Tollbit in the United States, explained how the company helps publishers monitor, manage, and monteise AI bot scraping and using their content.

“We give you visibility into what is actually happening out there: how many times are you getting scraped and by whom? Are you getting referrals back?” Stone said. “We allow publishers to charge for access on a usage-based model.” 

Madhav Chinnappa, vice president of partnerships at Human Native AI in the United Kingdom, offered an alternative approach: creating a marketplace where firms can buy licenses for content from publishers to train AI models.

“We’re trying to create a different economic ecosystem,” Chinnappa said.  

Regardless of which approach publishers decide to embrace, Hopperton had a clear message for attendees: crafting an effective AI strategy requires media companies to have an informed view of how AI agents consume their content.

“One thing I end up saying in every conversation about AI is if you do not already measure and or block bots, please do that,” Hopperton said. “At least measure it.”

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .