Roughly two years ago, I worked with a PublicSource reporter on a story about how college professors were grappling with ChatGPT, then a newcomer in the AI world. Some were trying to guard against it. Others saw it as just another thing to watch for in the cheating detection game. And one professor, who led our story, proposed introducing it to students — working with them openly to explore what the tool could do.
That acceptance of change — coupled with the sense of responsibility to usher it in the right way, and the need to illuminate people’s highest-value skills — stuck with me.
AI has been around for a while. But recently, we’re hearing about a new feature every day, and they’re more accessible than ever.
Many industries are trying to figure out how AI can and should be used, and journalism is not immune to that. We know adoption of new tools can be fraught. We’ve reported on Pittsburgh authors and publishers who feel ripped off by AI and how some labor unions are pushing for AI clauses in contracts to protect workers. And there are ethical implications for the environment as well.
At the same time, we can’t afford to bury our heads in the sand about emerging tech. Anyone who has followed the news industry knows that didn’t work out too well last time when newspapers tried to resist the internet.
So in true PublicSource fashion, we’ve been gathering thread over the past two years:
- engaging in research about AI in journalism,
- participating in trainings with The Associated Press and through Report for America,
- consulting with a top CUNY professor who is advising newsrooms around the country,
- bringing learnings back from conferences,
- gathering input from our team and board members,
- and observing what other mission-driven newsrooms are doing with AI.
Here at PublicSource, we see some AI tools as a way to support our journalists and team members in serving you better.
We don’t want AI to be used in the dark or without the proper training. And if it is to be used, we want to do it in a clear-eyed way that furthers our mission to inform and inspire the Pittsburgh region through the power of deep, independent journalism.
PublicSource now has an AI policy and guidelines as a foundation for exploring this new territory. The policy prominently emphasizes our values of human-centered journalism, transparency, accuracy and privacy protection. We’re establishing a tech committee to regularly evaluate proposed uses of AI and other technology that can help us do our jobs.
The policy also clearly states approved and prohibited uses of AI at present.
We will allow AI to assist with research, transcription, accessibility and summarization in some cases.
For example, we are already using an AI tool to help create alt text for photos; this is what screen readers use to describe images for users who are blind or visually impaired. We’ve also introduced occasional audio readings of our stories, powered by AI. We don’t have the staff capacity to create these recordings, but offering them increases accessibility for those who need or want to listen.
We may also explore AI assistance to help us show up on various online platforms in ways that are most likely to get to you because — let’s be honest — AI may be better at algorithm-ese than we are.
We won’t allow AI to write or edit articles, manipulate images or make editorial decisions. And we will continue to guard source confidentiality and data privacy.
Human review on any AI use is absolutely essential. You’ll never see a lick of AI that hasn’t been carefully shaped and scrutinized by organic intelligence (meaning the brains of our skilled journalists and other team members).
Our fact-checking practices, which exceed many industry standards, will continue to serve us well through this evolution, and we will continue to look for methods to bolster our efforts and safeguard against complacency.
We’ll also disclose to you if we substantively use AI in the practice of our journalism and invite you to share your feedback now and in the future.
There is no PublicSource without our people — your neighbors — spending time with you in your communities, your living rooms, on the phone, on social media, wherever it may be. That’s not changing. We hope only to harness new technology to serve you in more meaningful ways.
To learn more about our AI policy, visit our policies page here.
Halle Stockton is the editor-in-chief and co-executive director of PublicSource. Contact her at hstockton@publicsource.org.
