Press Gazette: how journalists are using AI in 2026
This Press Gazette piece by Guy Abramo, CEO of Cision, draws on the 2026 State of the Media report, which surveyed nearly 1,900 journalists across 19 markets. The article presents the data as a corrective to two opposing narratives: that AI has taken over journalism, and that journalists are ignoring it. The findings suggest something more specific — AI has become common, it is concentrated in background tasks, and journalists are still the ones deciding what matters.
Adoption is rising but uneven
21% of respondents say they don’t use AI, down from 33% a year earlier. In Asia-Pacific markets, the figure drops to 11%. The trend line is consistent: each survey cycle shows fewer non-users. What it doesn’t show is what adoption means in practice, which is where the detail is more interesting.
Most use is for background tasks
The most common uses are brainstorming angles and interview questions (48%), research and fact-checking (43%), and transcribing or summarizing interviews and audio (41%). These are preparation and processing tasks rather than drafting or publishing tasks. Only 15% of respondents said they use AI to write a first draft.
This distribution matches what researchers in adjacent fields find: professionals tend to adopt AI for work they don’t consider central to their identity. For journalists, that appears to be pre-reporting research and post-recording processing. The story itself, the judgment calls about what to pursue, verify, and foreground, remains human-led.
Journalists push back on AI-generated PR
53% of respondents said they oppose receiving AI-generated pitches or press releases. 25% are neutral and 21% are in favor. The objections are specific: marketing language, unsourced claims, and irrelevance top the list. One journalist quoted in the article warned against relying on AI platforms as sources because they can be significant vectors for misinformation.
This finding has practical implications for PR and communications teams. Journalists who have adopted AI for their own work are not necessarily more tolerant of AI-generated material arriving in their inboxes.
The profession’s main concern is accuracy
50% of respondents named accuracy, fact-checking, and countering misinformation as the top challenge facing journalism. 43% cited navigating AI’s rise and impact. These two concerns are related: AI tools can produce plausible-sounding errors at scale, which puts more pressure on the fact-checking work journalists already consider central.
The article is worth reading alongside newsroom policy documents and training materials. It provides a data-grounded picture of where journalists currently place AI in their workflow, which is useful context when setting expectations about what AI governance needs to address.