Generative AI and Audiences: Revisiting UK public attitudes to AI in media
Generative AI has entered the mainstream in the UK. Our latest research for the BBC shows awareness is near universal (98%) and a majority (58%) have used GenAI tools, with one in three (35%) now using them weekly. Usage is rising fastest among older groups (a 250% increase among 55+ year-olds), and families are heavy adopters (74% have used).
Yet deeper use has brought deeper questions: people value AI’s convenience but worry about over-reliance, authenticity, and trust, especially in media.
Media matters because it shapes culture, identity and public understanding. In this context, audiences are cautious: 58% say AI in media makes them nervous, 70% prefer human-driven movies, and 78% prefer human-written online news. The emerging rule is clear: GenAI is welcome when it helps behind the scenes; it is resisted when it replaces human creativity, voice or editorial judgement.
The Impact on Media
The study suggests strong acceptance for supportive tasks for audio, such as AI-generated playlists, background music, and audio versions of written articles. However, there is clear resistance to AI-hosted podcasts, cloned presenter voices, and AI voice-over for live sports commentary, where personality, spontaneity and emotion matter.
For visual and video, there is a growing comfort with AI-enhanced production (e.g., animation, VFX) as a natural evolution of creative tools. On the other hand, we saw pushback against AI-written scripts or fully AI-generated shows, suggesting audiences believe storytelling requires lived experience and human perspective.
Importantly, in news, tolerance remains narrow and functional. AI can assist with accessibility and headlines when a journalist leads the reporting. Importantly, audiences overwhelmingly reject AI writing original news or generating editorial imagery, citing misinformation risks and the erosion of trust.
A new benchmark for responsible use
In 2025, audiences apply three linked expectations:
Value: AI should be genuinely useful and enhance discovery, access or personalisation without diluting accuracy or authenticity.
Humanity: Protect the human elements of creativity, emotional nuance and judgement. Keep human oversight visible through bylines, editorial sign-off and clear responsibility.
Trust: Be transparent about where, why and how AI is used, and who is accountable. With half of UK adults believing AI will worsen online disinformation, verification matters; 43% say they are more likely to trust AI use in news if outputs are fact-checked. While 63% look to government and independent regulators for governance, they still hold media brands directly responsible.
Discover the full report here: https://www.bbc.co.uk/mediacentre/articles/an-update-on-ai-at-the-bbc
About the research
Ipsos conducted a 10-minute online survey of 2,000 UK adults aged 16–75 (April–May 2025), alongside six in-depth online workshops with a cross-section of participants across the UK. The findings track shifting behaviours and expectations and outline where audiences now draw the line on AI in media.
