The European Federation of Journalists (EFJ) has published a study based on surveys of 30 journalists' unions across 19 European countries. It shows journalists are heavy users of AI for transcription and translation, but they are deeply worried about deepfakes, copyright, and editorial independence. Most are now demanding training, legal guidance, and collective bargaining rules to keep the technology in check.
The study, "Study on AI and work in the media sector: journalists want more training," found that 76% of journalists use AI for transcriptions and summaries of interviews or recordings, and 76% use it for automatic translation and subtitling. But the same journalists see AI as a direct threat to their craft's core values.
"Journalism's concern profile is dominated by the public-trust axis. Deepfakes/misinformation (93.5 per cent) is the defining peak, closely followed by misuse of AI-generated content (90.3 per cent) and ethical concerns around automated reporting (90.3 per cent)," the report said. "This indicates journalists see AI as a threat to verification, legitimacy, and accountability. These are risks that extend beyond the newsroom to audiences and democratic discourse."
The survey also captured union perspectives. Journalists' unions reported consistently high levels of concern across all surveyed risk areas. The highest "very concerned" scores clustered around intellectual property rights (60%), deepfakes and misinformation (60%), and misuse of AI-generated content (52%). Remuneration and working conditions also ranked highly, with 72% concerned or very concerned.
The report criticizes the media sector's response so far. "A recurrent theme is that the media sector is reacting without overall direction or policy: There is no strategy or workflow regarding AI, only ad hoc press statements and going with the flow type of thinking," the report said.
What journalists want
The demand for safeguards is high. Nearly two-thirds of journalists want training or workshops on AI tools (64.5%). The report's authors argue that decision makers underestimate the urgency of AI literacy among journalists and audiences. For writers seeking the training the survey found to be so widely demanded, resources like AI for Writers and AI for Creatives cover the necessary basics.
Large majorities also want structural protections. About 87.5% want clear, accessible information on how AI is used in media, and the same percentage requests legal guidance on copyright, data use, and algorithmic transparency. Three-quarters demand collective bargaining frameworks, and 75% call for training programmes and monitoring tools.
They warned that a growing reliance on AI has led to faster but often shallower content, with less context and fewer verified sources. "These reports do not claim to provide a final or exhaustive account. AI is developing rapidly, and the perceptions captured here reflect a specific moment," the report concludes, framing its findings as a baseline for policy responses and social dialogue.
Why this matters for writers
For writers, the EFJ data is a negotiating tool. It shows that employers are already integrating AI without a clear strategy, which puts the burden on individual writers to protect their rights. Treat AI policies like any other workplace safety issue: demand transparency from your employer on where AI is used, secure your copyright in AI-assisted workflows, and document your original input.
Your membership also unlocks:
