Can AI-Powered News Video Ever Be Trusted?
Trust in news video depends less on the label and more on what the technology was actually asked to do.
Becky FoxDirector of Marketing, Genna6 min readSubscribe on LinkedInThe words “AI” and “news” strung together probably aren't something everyone wants to hear these days. AI is in everything nowadays and there are good reasons to be cautious about the role it plays in journalism.
But “AI in news” can mean wildly different things.
On one hand, AI can help an editor transform a finished article into another format. It can transcribe an hour-long interview and help find the one great soundbite you remember by only a handful of words. Hello, viral clip.
On the other hand, AI can generate an image of something that never happened. By most editorial standards, that's a “no go” to the highest degree.
Those uses shouldn't be treated as the same thing. And audiences appear to recognize the difference. The Reuters Institute's 2025 Digital News Report found that people across markets expect AI to make news cheaper to produce and more up to date, but also less transparent, less accurate and, most notably, less trustworthy. Audiences are also more comfortable when humans remain involved: separate Reuters Institute research found just 12% were comfortable with news made entirely by AI, compared with 43% when a human led the process with some AI assistance.
So whether AI-powered news can be trusted depends a lot less on the label “AI-powered” and a lot more on what the AI was actually asked to do.
AI-Powered Doesn't Have to Mean AI-Generated
It's becoming commonplace to put everything involving AI into one giant bucket.
AI summarized a script? AI news.
AI helped edit an hour-long press conference? AI news.
AI created a fake meteor crashing into Midtown Manhattan? Also AI news.
There is an important difference between AI generating an untrue reality and AI powering production processes around real content. For publishers and newsrooms, that distinction starts with something they already have: a reported article, an interview, an approved script, an RSS feed or an existing media archive.
Instead of asking AI to invent the story, AI-powered video production can help transform material that has already gone through an editorial process into something useful for another format.
A published article can become a video for your website or social channels. A recorded interview can become several shorter clips. Weather data can power recurring local video. The video doesn't need to start with AI. It can, and should, start with journalism.
Trust Starts with What Goes In
Before worrying that your local Chicago news video is going to show The Bean growing hands and feet and walking down the Magnificent Mile, look at the source. If a newsroom has already reported, edited, and published a story, that's fundamentally different from asking AI to “make me a video about what happened in Chicago today.”
That distinction matters because trustworthy newsroom automation still needs a trustworthy foundation. Publishers should be able to trace what they're creating back to source material they understand and stand behind. If the original reporting is the foundation, AI can support production without becoming the reporter itself.
It’s true, technology can help with repetitive production work while journalists retain responsibility for the journalism.
And What About the Visuals?
If your story is about a real flood happening in a local neighborhood, the audience probably shouldn't have to wonder whether the flood they're looking at ever existed. Trust isn't only about the words in a story, it’s also about the visual evidence accompanying them.
Where did the footage come from? Does it actually depict the event being discussed? Is it owned by the publisher or properly licensed? Was any part of it artificially created or altered?
Research suggests audiences are thinking about those questions, too.
In a 2025 Pew Research Center study exploring what Americans consider to be “news,” participants often paid more attention to content that included visual evidence, and eyewitness or on-the-ground footage was seen as more trustworthy. At the same time, several participants raised concerns that AI is making it harder to trust what they see. That tension is worth paying attention to.
AI can help identify, organize and assemble media without needing to manufacture a visual version of reality. That could mean working with your own footage and archives, media you already have rights to use, or licensed editorial sources. The technology around the video can change considerably without changing the expectation that what audiences see should accurately represent the story being told.
For newsrooms, that makes the distinction between real-world visuals and synthetic imagery more than a product feature. It's part of maintaining editorial trust.
Editorial Control Has to Survive Automation
“Human in the loop” has become one of those phrases that appears in almost every conversation about responsible AI. But simply having a human somewhere in the process doesn't tell us very much.
The more useful question is: What can that person actually control?
- Can an editor rewrite the script?
- Can they reject or replace a visual?
- Can they change the edit, captions, narration or branding?
- Can they stop the video from being published?
- And, ultimately, who has final approval?
AI-powered production can handle a growing amount of repetitive work. It can help structure content for video, transcribe footage, surface relevant moments, organize assets and prepare content for different formats.
But efficiency shouldn't make editorial judgment disappear. AI can help make the first draft. The newsroom should still get the final word.
Trust the Workflow, Not the Buzzword
At Genna, this distinction shapes how we think about AI-powered video. Our tools help publishers transform existing journalism, structured data and media into video while keeping editorial teams in control and using real-world, non-synthetic imagery. But the principle is bigger than any one platform.
AI-powered news can be trustworthy. It can also be incredibly untrustworthy.
The difference comes down to what went in, what the technology was allowed to do, and who remained responsible for what came out.
AI should make it easier for journalism to reach more people. It shouldn't make audiences question whether the journalism they're looking at is real.
Subscribe on LinkedInReady to make video your newsroom can stand behind?
See how Genna helps teams turn existing journalism into publish-ready video with full editorial control.
Explore Genna solutions