Let AI help, but don't let AI tell the story

Many organisations now use AI to draft press releases, summarise reports or make illustrations. The question we hear most often in training sessions is: “Can we use it, and do we have to tell readers?”

The latest data does not answer yes or no. It points to a clearer line. People are not worried that AI is part of the process. They are worried that AI is about to replace the person who tells them the story.

The more AI does, the less comfortable people are

The Reuters Institute's Generative AI and News Report 2025, with YouGov, surveyed 12,000 internet users in six countries: the United States, the United Kingdom, Japan, France, Denmark and Argentina. It asked how comfortable they were with news produced with different levels of AI involvement [1].

For news written entirely by journalists, 62% were comfortable. When journalists did most of the writing with some AI help, that fell to 43%. When AI did most of the writing with a human checking, it was 21%, and for news written entirely by AI, only 12% [1].

Notice that even with a human as the main writer, simply adding AI help cut comfort by nearly a third.

Backstage work is fine; the public face is not

Broken down by task, the picture gets sharper. Before you look at the answers, pick the tasks you would be happy for AI to do, then compare with people in the six countries [2].

Most people were most comfortable with AI fixing spelling and grammar (55%) and translating (53%), and least comfortable with realistic images used in place of photos (26%) and with AI presenters or authors (19%) [2].

This fits what we saw in our previous article: people trust people more than channels. When AI does work readers cannot see, such as checking spelling or analysing data, they hardly mind. When AI becomes the face or the voice speaking to them directly, trust drains away quickly.

Why this matters especially for Thai organisations

This survey did not include Thailand, so we cannot say that Thais think the same way. There are still reasons for Thai organisations to be more careful than before.

In 2026, overall trust in news in Thailand fell to 47%, the lowest in six years of data [3]. When trust is already thin, using AI where readers feel a machine has replaced a person could wear away what little is left.

On the other hand, across the six countries surveyed, 37% believe the news media already use AI often, and only 4% think they never do [4]. Readers are not naive. Many already assume AI is at work behind the scenes, so hiding it is riskier than saying so plainly.

Three principles for using AI in communication without losing trust

1. 1

   Use AI fully behind the scenes

   Checking spelling, translating, summarising documents, analysing data and producing first drafts save a great deal of time, and readers accept them fairly well.

2. 2

   Keep the face and the voice human

   Speakers, signatories and images presented as real events should be real people and real photographs. If you need an AI-generated image, use it as a conceptual illustration, not as something that looks like a news photo.

3. 3

   Tell readers plainly

   Say briefly which part AI helped with and who checked it and takes responsibility, for example “First draft prepared with AI; reviewed and approved by the corporate communications team.” This tells readers that a person is still accountable.

Reading this data with care

The main data in this article come from an online survey in six countries in mid-2025, which did not include Thailand. Attitudes to AI may differ with culture and familiarity with technology, so treat the figures as a signal of direction, not as a stand-in for Thai audiences. If your organisation needs to make a policy decision, surveying your own audience will give a more accurate answer.

AI can help, but trust still has to come from people

AI really does make communication work faster and broader. But the data tells us that people still want to know who is behind the stories they hear. Organisations that use AI wisely let it do the heavy lifting backstage and put trusted people out front.

If your organisation is drawing up guidelines for using AI in communication, the Center of Excellence in Communication Innovation at Chulalongkorn University's Faculty of Communication Arts would be glad to talk and share what we have learned.