Fourteen UK news organisations told the National Council for the Training of Journalists that graduates need more than a basic familiarity with artificial intelligence, Journalism UK reports.
The skills employers highlighted include critical prompting, fact-checking, context awareness and ethical judgement — especially knowing how to identify bias, verify AI-generated material and spot hallucinations, according to Journalism.co.uk’s account of what recruiters are asking of new journalism graduates.
That matters because the NCTJ remains a major route into British newsrooms. Its Journalists at Work 2024 survey found that 81% of UK journalists hold a journalism qualification, and 83% of those qualified journalists went through NCTJ programmes.
The training body is now changing how AI is taught and assessed. It says AI literacy will be integrated across existing parts of the Diploma in Journalism rather than treated as a standalone subject. AI will also appear in exams and portfolio work, with students expected to disclose and reflect on how they used AI tools.
The research behind those changes is relatively small. The NCTJ describes it as qualitative research based on discussions with 14 journalists and editors from national, regional, specialist and broadcast news organisations. The published executive summary does not identify the organisations or break the findings down by specific newsroom role.
Still, the pattern was consistent in one respect: everyone consulted was using AI in some form. Some organisations were moving ahead more aggressively, while most described their approach as “cautious innovation.”
The NCTJ also found widespread shortages in AI literacy inside newsrooms. That echoes a separate survey of 448 newsroom leaders across 86 countries, which found that skills gaps and staff skepticism are among the biggest barriers to wider AI adoption. Six in 10 respondents said their newsrooms offered no formal AI training.
For journalism students, the message is less about mastering a particular chatbot than about judgment. The NCTJ says training should emphasize when AI is appropriate, how to verify its output, how to recognize hallucinations and when human editorial control is essential.
Not every student is eager to embrace the technology. A Northeastern University ethics seminar found journalism students pushing back on AI more strongly than their professor expected. But that classroom experiment was small and American; it is better read as an example of the debate journalism schools are navigating than as evidence about UK graduates broadly.
The changes also arrive during a difficult period for newsroom employment. More than 2,300 journalism jobs were lost in the U.S. and U.K. in the first half of 2026, according to a Press Gazette tally we previously covered. But the NCTJ’s own research cautions against drawing a simple line between AI and those cuts. None of the people it consulted said AI had directly reduced the number of journalists in their newsroom at that stage.
That distinction is central to the NCTJ’s approach. It is not asking students to replace reporting skills with AI skills. Reporting, verification, media law, ethics and editorial judgment remain the foundation.
The new expectation is that graduates understand how AI fits around those skills — and, just as importantly, when it should not.
