To ponder the injection of artificial intelligence into journalism is not a source of joy for this local reporter.
When ChatGPT was released, I basically ignored it. To this day, I try my best to pretend generative AI doesn’t exist – until of course, I try to Google anything, and it slips past my self-imposed blinders. Or until I’m looking up something related to my previous reporting, and an AI-generated summary of my own article appears at the top of the page.
I think local journalism is a relationship between readers and reporters that’s based on trust – that the reporter asked thoughtful questions of real people, vetted research, and distilled the received information themselves as truthfully and accurately as they know how. And my big worry (among several others) about AI’s steady foray into global journalism, as both a journalist and consumer of news, is that the information could be increasingly wrong.
I reflected on this while attending the International Symposium on Journalism at UT-Austin, which took place on Sept. 17 and 18, sitting in a workshop for reporters focused on the “AI tools” we could use in the newsroom. “Chat with Gemini to research, plan, and explore story ideas and more with Google AI,” the screen at the front of the room read.
What I knew going into the conference is that many newsrooms, no matter how resistant to AI, are needing to publicly grapple with how they’re using it for transparency’s sake. The Chronicle published its own AI policy earlier this month. Another reason for such policies: A reader asked me several weeks ago if I use AI to write stories, and was pretty shocked (to my great offense) to learn that I do not. “More people should probably know that,” he followed.
What the conference forced me to begrudgingly confront, especially as an early-career reporter, is that a lot of people are increasingly consuming news that’s created by social media influencers and AI-generated sources – and even trust those sources over articles written by trained journalists. “We are at a moment where content is basically becoming infinite,” Jonathan Greenberger, global editor-in-chief at Politico, said Friday.
Zach Seward of The New York Times said the outlet regularly studies their readers. “And lately, when we speak to them, a lot express the view that one: Journalists are inherently biased, like all humans. And two: that AI overviews, because they combine information from a range of sources, are more objective and trustworthy than traditional news,” Seward said. “Trust in news has been declining for some time. We have to acknowledge that’s part of the dynamic.”
They asked us reporters to keep pace. “Any journalist who would refuse on principle to use AI tools to help them be an even better journalist – I don’t understand that,” Greenberger said.
Another reality is that our reporting is being masticated by machine learning models and haphazardly spat back out, often without much credit or click-back to the newsroom that produced that reporting. On Friday, newsroom leaders exchanged ideas for how to monetize that transaction and charge AI agents to use their content, which seems to me a rather depressing agreement.
As an alternative, hundreds of newsrooms are challenging AI’s free absorption of their work. The Chronicle has now joined more than 550 other media outlets in suing OpenAI and Microsoft for pirating our reporting at will. (You can read more about the copyright lawsuit in Austin Chronicle Publisher Cassidy Frazier’s column.)
I hear the arguments for AI’s benefit. As someone who often opens massive government spreadsheets and thousand-page PDFs, AI could really help comb through the overwhelming amount of information. I hit the Gemini button on Gmail to put interviews on my schedule. I use Otter.ai to transcribe my interviews, which is a lot faster than listening back on hundreds of hours of audio and more accurate than my furious scribbling in a notebook.
And in the many positive examples of AI use that newsroom leaders brought up over the course of the conference, they pressed that human verification is a necessary part of the process.
That sounds good in theory, and yet I have lingering concern that the human error that already naturally occurs in the reporting process (that usually gets caught by the fact-checker) could only become more common with AI in the mix. And because whenever the little star and AI-generated summaries have appeared at the top of my favorite outlets’ articles, I felt a new subconscious cause to question the information.
Moreover, I have no interest in being an influencer. I’m a writer for a print publication and would like to stay that way. This early-career journalist can only hope that the pendulum of public trust – no matter how polarized our political climate – always swings back to human work, and that deeply and distinctly reported stories carry value of their own. That, as Seward said Friday, “People enjoy a good yarn.”
This article appears in September 25 • 2026.
