AI Trust Kit helps newsrooms win back skeptical readers

Lynn Walsh, assistant director, Trusting News (Photo credit: Trusting News)

We might need to disclose things now that will become so standard and sort of best practices that we won’t need to disclose them in the future. As people learn more about this and are confused by it or skeptical of it, their expectations and needs for disclosure might change as well. We’re constantly learning as much as we can about people’s frustrations and confusion around journalism.”

Joy Mayer

If trust is a two-way street, then newsrooms will need to go the extra mile with transparency when implementing the use of artificial intelligence (AI). Nonprofit Trusting News recently put together a research-based AI Trust Kit, advising reporters on best practices when considering how to employ the emergent technology, which fewer than 32% of Americans surveyed say they trust.  

“Currently, I’ve been focusing on our AI technology and product work, and what that means is we want more people to have access to responsible, ethical information,” said Lynn Walsh, assistant director at Trusting News. “We know that journalists, for the most part, are still a main source for that information. We’re trying to help them build trust with the public so more people consume their content.”

True to the name, Trusting News researches how journalists can rebuild and maintain public confidence. The nonprofit dates to 2016, when founder Joy Mayer left her job teaching at the University of Missouri’s School of Journalism after becoming increasingly frustrated with the public’s lack of faith in “the basic integrity of journalism,” fueled by a fundamental misunderstanding of the practice.

“So, I wanted to learn as much as I could about how trust is gained and lost, and what journalists could do in the face of lower trust and accusations,” said Mayer. “It started as a part-time fellowship at the Reynolds Journalism Institute at the Missouri School of Journalism, and just grew from there into a project that, for a few years, was based at Mizzou. Then we were co-hosted [for] several years by the American Press Institute and the Reynolds Journalism Institute. In 2024, we spun off to be our own nonprofit.”

Within the AI Trust Kit are materials, such as tip sheets, sample questions and real-world examples, to help newsrooms properly develop their AI use policies and remain mindful of best practices. But what are the best practices? The Trust Kit recommends finding out through audience engagement and provides four tools, including a survey template, to solicit community feedback.

Walsh said newsrooms should be asking readers how they feel about journalists using AI and what they expect when journalists do use AI. “The reason we recommend starting with talking to your community is that we have found that there’s a broad range of what people are comfortable with and not comfortable with,” she said.

In fact, Walsh, a former national president and ethics chair of the Society of Professional Journalists, began her own research for the guide through outreach to newsroom partners. Respondents generally found using AI for transcription and checking grammar and spelling acceptable. However, they felt uncomfortable with AI writing stories, even with human review.

The Trust Kit also addresses how newsrooms can disclose AI use, with a guide on drafting an AI use policy and a template for disclosing use in-story. Walsh’s research points to readers particularly caring about why journalists decided to employ AI in their reporting process and what they did to ensure their use of AI was ethical and accurate, along with whether human review was present.

The Trust Kit also suggests newsrooms can play a role in directly educating audiences about AI. “Here’s an opportunity to say, ‘Let us help you learn along with us,’” said Walsh. “’Let us help you understand how this is impacting your kids, your life, your schools, your government and the environment.’”

She said the education piece remains a working document for how reporters can help demystify AI, with more tools to come. In fact, Mayer added that the Trust Kit is already one of the most adapted since publication, and she expects it to continue evolving as the technology evolves.

“We might need to disclose things now that will become so standard and sort of best practices that we won’t need to disclose them in the future,” said Mayer. “As people learn more about this and are confused by it or skeptical of it, their expectations and needs for disclosure might change as well. We’re constantly learning as much as we can about people’s frustrations and confusion around journalism.

“Trying to help journalists realize that they can be frustrated that people don’t trust them, or wish that people trusted them. But that doesn’t do much to change the situation. Every journalist needs to figure out what they can do today to demonstrate that their work is worthy of trust.”

Tandy Lau is an award-winning journalist currently based in Jersey City. He currently works as a staff reporter covering public safety for the New York Amsterdam News in Harlem. He can be contacted at tandylau1995@gmail.com.