Case Study

Lenfest AI Fellow spotlight: Omar Reid

The Boston Globe

October 19, 2026

This profile is part of the Lenfest AI Collaborative and Fellowship Program’s 2026 Impact Report.

As an AI Fellow at The Boston Globe, Omar Reid is tackling brand safety, personalization, and data silos to help the newsroom unlock the full value of its journalism.

About his role and the problem he is trying to solve

Omar’s path into journalism is unconventional but tightly focused on AI.

“All of my career, I’ve basically been an AI engineer,” he explains. He worked briefly at a bank, where he realized that he was merely a cog in a huge machine and unlikely to have an impact. So, when he had the opportunity to do a one-year work placement, he went to a startup in Tokyo, Cogent Labs, which had an English website and had hired many foreign engineers.

At The Boston Globe, Omar’s role sits at the intersection of data, AI, and organizational change.

The core challenge he’s working on is what he calls a “family of problems” rooted in silos:

Data silos: “All the data sets [are] in separate systems.”
Organizational silos

To tackle this, he centralized archives and built an internal AI chatbot in Slack that lets colleagues search across the Globe’s data.

“Once they saw that, they were like, ‘Wait, hold on, we have all of this useful data, and now we have… a way to access it,’” he said. That proof of concept helped build trust.

Now, he’s applying that same approach to a high‑stakes problem: brand safety and the long-term opportunity of personalization.

His project at The Boston Globe is built around extracting rich signals from every article — topics, people, places, organizations, and sentiment — so the ad system can understand what a piece is actually about, not just which section it sits in. It is designed to give advertisers fine‑grained control to avoid sensitive content while targeting context that aligns with their brand.

Longer term, the same signals that power brand safety can also underpin personalized products such as tailored newsletters, letting the Globe surface stories that match what individual readers actually care about while protecting advertisers from risky contexts.

The biggest surprise during the fellowship

Omar said his biggest surprise has been the disconnect between the influence of news organizations and the money they make, especially compared to tech giants.

“I didn’t realize… news used to run things, have a lot of power,” he reflects. When he joined the Globe, he even told them in his interview, “Oh, the Globe still exists?”

Over time, he came to understand that the Globe is “still around and kicking,” with real cultural presence and prestige, but operating at a very different financial scale from tech platforms.

“You expect it to be making Meta money, but Meta is like a trillion-dollar company… It’s like a really small group of people in this industry.”

Another surprise has been how much value is locked up in news archives. His own work is driven by that realization: “I’m less worried about the views and clicks, because I know that they will happen if I put my attention on surfacing the value from all of this knowledge that’s been accumulated.”

One thing the industry gets wrong about AI

For Omar, the problem is skepticism amplified by misconceptions.

“Reporters seem to be very skeptical, which I think is a good thing,” he said. But when that skepticism is “married with the misconception,” it can become paralyzing.

He describes speaking with people who think “AI started in 2023 when ChatGPT was released,” and who rate the danger level of AI on a scale of zero to 10 as “10 – Terminator.”

“Whereas to me, AI is like a clever washing machine,” he said. Through conversations, he has tried to bring people back from science‑fiction fears to a more realistic view.

“It just makes it hard to do useful stuff” when the conversation is dominated by apocalyptic narratives, he pointed out.

Lessons that other newsrooms could apply tomorrow

Omar’s advice is simple but difficult: Slow down enough to learn.

He sees many journalists and newsrooms as “conditioned to react,” driven by constant urgency.

That same mindset, he argues, undermines meaningful experimentation with AI: “If I tell you now to sit down on a computer, spend 30 to 40 minutes playing with something… this is just discovery now. This is not you reacting to anything. That becomes very hard if you’re very urgent and reactive.”

Newsrooms need to consciously drop the urgency when approaching AI.

“Be patient with yourself, be patient with the process, be patient with the computer,” he said, to build real intuition about where AI can help and what’s just hype.

What he is paying attention to right now

Omar is focused less on shiny new objects and more on building the right things.

He’s been reading “Continuous Discovery Habits” by Teresa Torres, a book that “outlined the process which people can use to know what is worth building.” With AI making it “really easy to build things,” he worries that teams will “start building things that in the end no one wants, because they didn’t think to check first.”

He’s also spending time with the revenue team to deeply understand the Globe’s business model. “What are the core levers here? And they tell me what it is, and I’m like, that is perfect. Now I know what we should target and why it will move the needle and produce revenue for the company.”

This lens helps him avoid “solutions in search of problems” and instead align AI projects with both user value and business value.

What he is most optimistic about

Omar believes we are at a turning point for language‑driven organizations, after years of watching numbers-driven organizations succeed.

“We deal with language, and now we have a tool that makes it really easy for us to deal with language,” he said, framing it as journalism’s “Moneyball moment” — a time when the tools finally match the domain, just as data analytics transformed baseball.

The Fast Three

A book, article, or newsletter you’d recommend?
Continuous Discovery Habits by Teresa Torres — “the number one thing people need to discern what’s worth building and not worth building.”

What’s one thing you’re learning that has nothing to do with AI?
How to deal more effectively with issues of identity.

One word that describes the next year of AI in news?
“Consequences. If you successfully understood the hype, you know what was hype, what was real, and you put the work in, looking at the problems in your company, trying to solve those, and you use AI to help you do that, you’re going to reap what you sow.”