I am experimenting with the use of Artificial Intelligence in my reporting on this website. My goal is to make clear when and how I use AI so that you can better judge the credibility of my work.
When writing freelance articles for outside publications, I will limit my use of AI, reverting to the kind of shoe-leather reporting I learned before all these machines and screens came into our lives.
To learn how journalists should best use AI, I rely on such organizations as the Global Investigative Journalism Network and the Pulitzer Center. The two organizations created an AI Accountability Reporting Guide which contains valuable resources and best practices.
I taught college-level journalism for 18 years in Florida. During my last few years in the classroom, I noticed more and more students using AI. Old-fashioned journalist that I am, I was dead set against it. I figured, how are students going to learn to write and think critically if they let machines do all their work?
My view on that hasn’t changed. I don’t think journalists should let machines write their stories and they certainly shouldn’t pass off AI text as their own.
But I have found that AI can be useful in:
- Quickly making sense of spreadsheets and data sets
- Getting up to speed on unfamiliar topics
- Creating infographics to illustrate complex ideas
- Transcribing interviews
AI can make mistakes, of course, so it’s important to try to verify and confirm results.
Over the summer, I used AI to analyze public works documents in Panama. That allowed me to quickly understand why some projects were delayed or over budget.
In another experiment, I took extensive measurements and drone photographs of sections of the Chiriquí Viejo River in western Panama to try to create detailed AI-driven flood scenarios. I discovered that bad measurements can ruin any predictions (See “Better data softens Chiriquí Viejo forecast, but an Eta-level storm still spells peril“).
AI can’t possibly help if you feed it bad information. Even asking a question with a certain tone can influence the results of a reply.
I have used AI to summarize documents, but I have found that such an imprecise approach can lead to mistakes and must be used with caution.
AI also has an inherent bias and admits it:
I am a reflection of the global information ecosystem. Recognizing that the ecosystem itself is unequal is the first step toward reading any analysis—including mine—with a critical, analytical eye.
See story: “How strange: AI admits bias against Cuban government.”
As I continue to learn how to best use AI, I will update this page.
What do you think? How much AI, if any, should be used in journalism? Feel free to leave your comments below and rest assured that human judgment – with all its faults – will remain at the heart of the stories on these pages.
Last updated in September 2026