Medium: AI is changing journalism — what the data says in 2026

Published in early June 2026 on Medium, this article by Techrefreshing compiles data from Muck Rack, the Reuters Institute, and synthetic media detection firms to describe how AI has actually changed journalism practice and business models — as opposed to how it was expected to change them.

Adoption data

According to Muck Rack’s 2026 journalist survey of approximately 900 respondents, 82% of journalists use AI tools regularly. ChatGPT leads at 47% adoption, followed by Google Gemini at 22%. Among news audiences, weekly use of AI tools nearly doubled year-on-year — from 18% to 34%, per Reuters Institute data. 65% of journalists surveyed still report finding their work deeply meaningful.

Structural changes

The article identifies the core tension: AI has improved certain parts of journalism while threatening its business model from a different direction. On the positive side, document analysis that previously required months of reporter time can now take hours. Data journalism and fact-checking workflows have been accelerated. Routine reporting tasks can be automated, freeing staff for more demanding work.

On the business side, AI answer engines that provide direct responses without linking to source articles are projected to reduce publisher traffic by 43%, per figures cited in the piece. Only 20% of publishers surveyed believe AI licensing deals will generate meaningful revenue for their organizations.

Synthetic media threat

The article flags a significant trend: synthetic media detection firm data shows 3,165 deepfake incidents recorded in March 2026 alone, compared to four in January 2020. The article uses Ireland’s 2025 presidential election as a concrete example — a deepfake video falsely showing the front-runner announcing his withdrawal circulated for days before being debunked. Smaller newsrooms face disproportionate verification burdens because authentication infrastructure requires resources they often lack.

Who this is useful for

Journalists and editors who want a current, data-backed picture of where AI adoption stands across the industry — specifically for conversations with management about tooling, hiring, or workflow changes — and anyone tracking the business-model disruption in news media through data rather than speculation.