How Boston Globe Media is putting AI to work in its newsrooms

A podcast recap: Building for Others with Shira Center and Sidhi Dhanda from Boston Globe Media

On this episode of Building for Others, we talk with Shira Center (VP, Innovation and Strategic Initiatives) and Sidhi Dhanda (AI intern) about four AI tools they built on elvex. Together they’re starting to change how a 150-year-old media company works.

Four use cases that actually work

School board monitoring at scale

Sidhi built a system that keeps an eye on school board meetings in nearly 100 Massachusetts districts. The setup is simple. Meeting transcripts go into Google Drive, an elvex assistant is connected to that folder, and journalists ask it questions in Slack.

School committee meetings run long, most of the agenda is routine, and plenty of them never produce a story. Scanning dozens of towns at once, though, makes it easier to spot themes and early signals a reporter would otherwise miss. Sidhi used the tool to find towns that were discussing AI literacy for an editorial, which shows how wider monitoring can feed both reporting and commentary.

The second tool tackles a problem every media company has: finding the right material across decades of content spread over several systems. Reporters can now search roughly 25 years of archives in plain language, again from Slack.

They ask things like “What was the first public comment Michelle Wu made when running for city council?” or “What was the largest snowfall in the last 25 years, and how does it compare to recent storms?” For questions about Boston, this is often a quicker place to start than regular search, especially when a reporter needs to get oriented or check a fact before digging in.

Headline generation with style

The third tool helps write opinion headlines that follow the Globe’s style rules. A writer pastes an article into Slack and gets several headline options that follow the paper’s rules on capitalization, structure and tone, with clickbait ruled out.

Staff mostly treat the suggestions as raw material. They take a phrase from one and a structure from another to reach a final headline faster, and the editorial call stays with them.

Alt text that people use

The last tool is the least exciting and one of the most used: an alt text generator for photos. Drop an image into Slack and it returns alt text formatted to accessibility guidance, which also helps with SEO.

It doesn’t change the journalism. It takes the pain out of a chore that has to get done, and that’s why people use it.

Why these tools actually get adopted

What the four tools share is where they live. They run in Slack, where Globe staff already spend their day, so there are no new logins, very little change to how people work, and less need for formal training. As Shira puts it: “The best way to introduce these tools is to build on other tools, and meet people where they already are. It’s habit stacking, plain and simple.”

AI’s role in media: a four-part framework

Shira sees AI paying off in four areas:

  • Newsroom tools: faster document review and first-pass synthesis.
  • Reader experience: using data to understand subscribers and serve them better.
  • Upskilling: helping staff learn the skills that will count as the tools change.
  • Business considerations: working through licensing, competition and how efficiently the work gets done.

Governance comes up again and again. The Globe moves fast where it can, but it protects the thing the business depends on, which is trust. “A human has put this content out there and has reviewed it,” Shira notes. “That is still incredibly important to us.”

Advice for builders: the learning curve is flat

Shira’s best line in the episode: “The learning curve is flat.” If you feel behind on AI, you can catch up fast. The fundamentals haven’t moved, and learning new tools is quicker than it has ever been.

Sidhi’s advice is hands-on. Make it easy to start by building where people already work. Try lots of use cases in different parts of the business. Keep iterating, and fail fast so you find what works. Show your colleagues what’s possible, because a demo does more than a training session.

None of this depends on having the most advanced model. The Globe picks specific problems, builds the tools into the daily workflow, and lets usefulness drive adoption instead of mandates.

The bigger picture

Local media is still squeezed: fewer reporters, tighter budgets and a business model in flux. The Globe’s AI work doesn’t fix that. What it does is stretch limited capacity in a few places, so the newsroom can watch more sources, search its own archive faster and spend less time on routine production work.

The point is to make parts of the job easier so reporters and editors have more time for the work that needs their judgment.

Bring your agent. Bring your team.