The American Journalism Project’s new “field guide” vets AI vendors for local newsrooms

For many newsrooms, figuring out whether or not they want to bring generative AI tools into their editorial work is only the beginning of a conversation. For those that do, there are a litany of follow-up questions. One of those is: Which tool exactly should we use?

AI products targeting journalists and news publishers have flooded the market in recent years, and few organizations are willing to publicly endorse specific tools. A new field guide from the venture philanthropy firm the American Journalism Project (AJP) aims to cut through the noise, with practical information and reviews on AI vendors.

AJP is most well-known for providing grants to nonprofit local news organizations across the U.S. Starting in 2022, it launched a new AI and Product Studio, backed by a $5 million donation from OpenAI. For two years now, the Studio has been providing grants and coaching for local newsrooms to bolster their AI experimentation, including Sahan Journal, Spotlight PA, and The City.

The field guide pulls on learnings from these grantees, as well as phone and email interviews, product trials, and insights from experts at The Lenfest Institute, Big Local News, and Microsoft. Rather than making blanket recommendations on which tools to use, the guide is meant to “share what we’ve learned so far so you can make informed decisions.”

The first iteration of the guide includes several AI tools that transcribe and summarize local government meetings, including Local Lens, See Gov, and City Meetings NYC. I covered Chalkbeat’s experiments with Local Lens earlier this year, including how it had been used by a beat reporter in Michigan to listen in on meetings for several school board districts.

For all the tools in the guide, AJP has listed both pros and cons, what the tool does, and which AJP grantees have already used it. The guide explains that a tool like Hamlet goes a step further than other meeting transcription tools by synthesizing data on relevant real estate projects and the voting histories of commissioners. One of the cons: it’s only available in cities with especially active real estate development.

You can read the full “AI for Local Reporting” field guide here.