Embracing AI as ‘assistant’ opens door to better election coverage
Playbook walks community newsrooms through creation of AI-assisted election dashboard
This article is part of “AI is here,” a series from RJI highlighting AI innovation in journalism. Read more here.
Bay City News, a digital news service covering 13 counties in the San Francisco Bay Area, has developed an AI-assisted strategy for local election coverage that creates opportunities for deeper and more efficient coverage — especially for community news organizations with limited time and resources.
Download playbook
Compiled into a free online playbook with the support of the Reynolds Journalism Institute (RJI) at the Missouri School of Journalism, the strategy involves using a data scraper to pull information into a central dashboard that provides voters with key information before and after local elections. The result is a prime example of “AI as assistant,” allowing reporters and editors to focus on impactful community coverage while automating the mundane but vital task of data collection.
“If you look at what the national organizations do, they are largely focused on the higher ticket races, and nobody goes down to the county and municipal level to help explain what’s on the ballot and actually cover results on election night,” said Katherine Ann Rowlands, owner and publisher of Bay City News. “So for a lot of smaller newsrooms, it’s a huge undertaking with literally hundreds of hours of dedication to making that happen. But it’s really hard to do it efficiently.”
The goal was to give Bay City News — and other news organizations across the country that can learn from its example — the breathing room to analyze and interpret election results, interview the candidates and make sense of electoral impacts on a community scale. This might seem like a straightforward task, but it has become increasingly difficult as budgets and staffs have shrunk, local news deserts have widened and audiences have shifted to social media or outright news avoidance.
For Rowlands and Bay City News, the answer was twofold: an AI platform, Claude.ai, capable of talking a novice through the process of coding a dynamic election dashboard, and an intern-turned-staffer ready to make it happen.
“I don’t come with a data science background, so I was heavily dependent on the AI,” said Ciara Zavala, who began as an audience engagement intern last summer and is now the outlet’s impact manager. “It was a back-and-forth conversation with the AI of troubleshooting and getting code that I could paste into WordPress.”
“If you look at what the national organizations do, they are largely focused on the higher ticket races, and nobody goes down to the county and municipal level to help explain what’s on the ballot and actually cover results on election night. So for a lot of smaller newsrooms, it’s a huge undertaking with literally hundreds of hours of dedication to making that happen. But it’s really hard to do it efficiently.”
Katherine Ann Rowlands, owner and publisher of Bay City News
Anthropic’s Claude.ai, nicknamed “Claudia” by Zavala and her colleagues, turned coding from a foreign concept to a process of trial and error with a patient, automated teacher. This assistance also made room for Zavala to put time, energy and skill into turning a data-driven dashboard into a visually appealing and accessible Election Hub.
Rowlands said those quality-of-life factors, far from afterthoughts, are essential to that most fundamental role of journalism: meaningfully engaging audiences.
“The dashboard looks fantastic, it’s very friendly to a reader and it’s easy to use,” Rowlands said. “That’s important, because if you just look at the raw results that the counties are pulling in, it’s pretty dense and not always easily searchable. The user experience was top of mind for us.”
And while voters and news consumers were top of mind, a close second was the community news industry, for which the RJI playbook offers a step-by-step accounting of the development process from conceptualization to implementation.
“News organizations can take this playbook and build something that serves the needs of their own audiences,” said Randy Picht, executive director of RJI. “This is a real, tested strategy that has already made positive impacts in the Bay Area, and I look forward to seeing its influence spread.”
Rowlands said she hopes to partner with other newsrooms to apply the strategy to the entire state of California, and she concurred with Picht’s vision of newsrooms throughout the country adapting the model to serve their own local communities.
“If you replicate the workflow — even if the details might be different — you can apply this to other states,” Rowlands said. “We have a country full of small newsrooms that are struggling with the same exact problem of shouldering a heavy lift on election night. This can make it so much easier and enable all of those newsrooms to focus more on the journalism and less on the data entry.”
See the Bay City News Election Hub here and review the RJI playbook here. To stay up to date on other RJI projects related to AI innovation and supporting the community news industry, consider subscribing to the free RJI Weekend Newsletter.
Cite this article
Fitzgerald, Austin(2025, May 16). Embracing AI as ‘assistant’ opens door to better election coverage. Reynolds Journalism Institute. Retrieved from: https://rjionline.org/news/embracing-ai-as-assistant-opens-door-to-better-election-coverage/
Comments