Sign in or Join the community to continue
AI at Work in Local News: How Newsrooms are using AI today
Posted Oct 01, 2026 | Views 57
# Newsrooms
# Journalism
# Small Business
Share
Speakers
Kevin Delaney
Editor-In-Chief @ San Francisco Standard
Kevin Delaney is a journalist and media entrepreneur who co-founded Quartz and Charter and led them to acquisition, was managing editor and a senior writer at The Wall Street Journal, and was senior editor at The New York Times and The Information. He's now editor-in-chief of The San Francisco Standard and Charter, the leading future-of-work media and research company. Kevin is a member of the Council on Foreign Relations, board chair at the Internews independent media nonprofit, and a graduate of Yale University.
+ Read More
Kristen Muller
Executive Editor @ The LA Local
Kristen Muller is Executive Editor of The LA Local where she leads a network of community-centered newsrooms serving more than one million Angelenos. Previously, Muller was Chief Content Officer of Southern California Public Radio (LAist.com and 89.3), where she helped transform the organization from a legacy public radio station into a multimedia local newsroom. She brings more than two decades of newsroom experience and an award-winning track record of growing digital audiences, deepening community engagement and leading newsrooms through transformational change.
+ Read More
Neil Chase
CEO @ CalMatters
Neil Chase is CEO at CalMatters, the nonprofit newsroom covering California policy and politics. He was previously executive editor at The Mercury News and East Bay Times, where his team won the 2017 Pulitzer Prize for Breaking News Coverage. He worked as an editor at The San Francisco Examiner, The Arizona Republic, CBS MarketWatch and The New York Times and was an assistant professor at Northwestern University's Medill School of Journalism.
+ Read More
Olivia Smith
CEO @ Media Consulting Coaches
Olivia Smith is an Emmy Award-winning journalist, media consultant and AI specialist based in Los Angeles. She is a versatile content creator and storyteller with more than 20 years of experience in print, broadcast and digital journalism. Olivia is the CEO of Media Consulting Coaches, specializing in media training, public speaking, content creation and AI strategy. She is also an adjunct professor at the USC Annenberg School for Communication and Journalism.
Olivia spent nearly a decade at Disney, working in inaugural digital positions for ABC News, Good Morning America and ABC7, and has contributed to numerous outlets including CNN, 60 Minutes, Al Jazeera, NY1 News and the Columbia Journalism Review. She has also worked on the GenAI Content Engineering team at Meta, guiding model training and optimizing machine learning processes. Her expertise in AI overlaps with her love of storytelling. Olivia has created courses for USC on journalism and AI, and she often leads workshops and speaking engagements on AI tools and strategy. Her goal is to empower creators to use technology ethically, maximizing their impact even with limited resources. Olivia created and hosts the AI podcast Prompt Response. Find her on social media @LivNews.
+ Read More
SUMMARY
In this Forum conversation, our panel explored how local newsrooms can use AI to reduce repetitive work, strengthen original reporting, and better serve their communities while preserving human judgment. CalMatters shared how its Digital Democracy platform turns legislative data into reporting leads and personalized newsletters that help readers engage with their elected representatives. The San Francisco Standard described using AI for investigative document review, audience analytics, and app features such as personalization, audio, and mapped stories, with leaders modeling practical uses to encourage adoption. The LA Local highlighted translation and synthesizing community questions, while emphasizing transparency, source protection, and human review, particularly when serving vulnerable communities.
+ Read More
CONTENT & TRANSCRIPT
00:00:00:07 - 00:00:23:02
Olivia Smith
So let me introduce our panelists today, Neil Chase, please come on up. Is CEO of Cal Matters, California's nonprofit newsroom covering policy and politics. He previously led the Mercury News and East Bay times, whose team won a Pulitzer Prize for breaking news coverage. Neil brings both editorial and organizational perspectives to how AI can support a newsroom.
00:00:23:07 - 00:00:30:11
Olivia Smith
Thanks for being here. Thank you. We have Kevin Delaney.
00:00:30:13 - 00:00:52:12
Olivia Smith
Kevin is editor in chief of the San Francisco Standard and Charter. He co-founded Quartz and Charter and has led newsrooms and new media ventures through periods of change. Today, he is exploring how AI can serve journalists and readers. And then we have Kristin Mueller.
00:00:52:14 - 00:01:12:22
Olivia Smith
Kristen is executive editor of the LA local, a network of community centered newsrooms across LA. She previously helped lead LA shift into a multimedia local newsroom. Her work focuses on reaching communities and building participation in local news. Thank you all for being here.
00:01:13:00 - 00:01:30:21
Olivia Smith
So before I ask you any questions, I want to start by asking the audience a question. Who here has ever felt with a show of hands? Who here has ever felt burnt out working in news as a journalist or just in your job in general? I hear a lot of laughs. Good. If you didn't raise your hands. I was about to call you lying.
00:01:30:23 - 00:01:50:10
Olivia Smith
I was about to call you all hours. I know I often felt burnt out working as a journalist for many newsrooms over many years, and here's what I watched happen every time technology made one part of the job easier. The newsroom didn't give us the time back. It gave us another job. Cameras got small enough for one person to carry, so the reporter became the photographer.
00:01:50:13 - 00:02:13:15
Olivia Smith
Editing moved to a laptop, so the reporter became the editor. The internet replaced the research desk. So the reporter became the researcher and publishing went digital. So the reporter became the social team. I tell my USC students how lucky they are to be able to shoot everything on an iPhone, and it's beautiful. When I started out, I had to lug around a 50 pound camera, but the iPhone didn't lighten the load.
00:02:13:15 - 00:02:32:06
Olivia Smith
Okay, maybe it did, literally, but you know what I mean. It just met one person was expected to do everything with it. So here's the flip. Every tool I just named made a job easier for a person to do, so the person then had to do the job. Maybe AI is different. It can do the grunt work before you get there and you check what it said.
00:02:32:07 - 00:02:50:17
Olivia Smith
So what if this is the tool that finally gives the time back, the one that takes the tedious step, the transcription, the formatting, parts of the editing off our plates instead of adding to them? Here's the thing, though. That's not up to the tool. It's up to us. The newsroom decides where the save time goes, and it decides how the tools get used.
00:02:50:18 - 00:03:12:11
Olivia Smith
Used right? AI gives you hours back, used wrong. It costs you a source, a correction, or your credibility. So what's the right way to use AI? That's what I want to ask you all today from three people who are figuring it out in real time. So let's dive in. Neil, we'll start with you. Let's get concrete. What is AI actually doing in your newsroom today?
00:03:12:13 - 00:03:34:08
Neil Chase
I love your setup for that. That's awesome. AI is doing a lot of things right. And when you encounter people who in the newsroom who are nervous about AI, I worried about how you're using it. It's like, okay, but how did you just transcribe that interview you took or the last web search you did, right? I mean, AI is it's infused into everything we do with a computer in front of you, right?
00:03:34:09 - 00:03:55:16
Neil Chase
So, so on that level, that's the part where it's changing everybody's lives without them maybe even noticing or paying attention. Right? More specifically, we're using it from some things that you might not be surprised by, like crunching through large amounts of data much more quickly to find stories, to process, to get things ready. Doing research on large amounts of information, transcripts, things like that.
00:03:55:17 - 00:04:24:01
Neil Chase
The most fun thing we're doing, the most extravagant thing we're doing, is something we built. Actually, it was built starting more than ten years ago at Cal Poly in San Luis Obispo. It's a tool called digital democracy. And the idea was that by collecting all the data you can possibly imagine about the state government, the state legislatures, hearings, videos, every dollar given to them in donations, every form they have to file, every trip and gift they've gotten, put all that stuff together.
00:04:24:01 - 00:04:43:20
Neil Chase
You can start to find some good story ideas. They built it, put it out there for the entire state of California. Everybody to use. And guess what? Nobody used it because it's a web tool just sitting there. We came in and restarted it and added journalism to it and said, let's use AI. And by the way, this was built long before there were llms that were publicly known.
00:04:43:22 - 00:05:00:10
Neil Chase
Right. So there's there's 11 different kinds of AI in this product. One of them is ChatGPT, the other ten most people have never heard of. And some of them are hand-rolled by some very old engineers who are trying to figure out now how to use some of the newer tools, why younger people run circles around them, which is the fun part.
00:05:00:12 - 00:05:23:03
Neil Chase
But the idea here is that you take a body of data, you collect it all, you clean it, you make it all fit into one taxonomy, and then you analyze it and find story ideas. So we're generating these tip sheets that say in the legislative data, here's a great story idea about something your legislator did or a committee did or something that happened, giving those to our journalists and our newsroom and then journalists all over the state.
00:05:23:03 - 00:05:45:23
Neil Chase
And now we're starting to roll out this around the country to other newsrooms like us. So it's big and intimidating and exciting and usually fun.
Kevin Delaney
Thank you. Kevin, I think they're probably like 3 or 4 ways in which our newsroom is using AI. The first one is like gets to your initial your introduction, which is just using it to do things like more quickly.
00:05:45:23 - 00:06:08:09
Kevin Delaney
So we're publishing an investigation later this week. And one of the reporters used it to create a chronology like a timeline sidebar, which actually was just like dates extracted from the reporting and the article. And so we'll run that. And we're obviously like checking in everything, but it saved a bunch of time to actually have have AI kind of create that for us.
00:06:08:09 - 00:06:35:21
Kevin Delaney
So there's there's a group of things like that. We use it a ton in investigative reporting to go through documents. One of my favorite examples, one of our politics reporters got a big dump of documents and the mayor's calendar in response to a public records request, and he put them into an AI tool and said his prompt, I think, literally was, find me a juicy story.
00:06:35:23 - 00:06:58:17
Kevin Delaney
I love that. And and like in actually it found a good story for him. It turns out that the mayor had been meeting quietly with Laurene Powell Jobs about some efforts to rebrand San Francisco, and like that was in the calendar documents. So that's like maybe the second thing, like there's capability like research reporting. That's that's quite powerful.
00:06:58:18 - 00:07:19:19
Kevin Delaney
The third area, which I find like really exciting and new, is actually using it for coverage strategy. So my experience with newsroom analytics over the years is that you would say like, okay, we're you know, we're publishing this big package of stories. Do we publish it on Saturday or Sunday or Monday or Tuesday or like what time of day?
00:07:19:19 - 00:07:46:12
Kevin Delaney
And we're optimizing for subscriptions. Like what should we think about. And and in fact, what we do is actually we take our analytics team created a Google Sheet that we can basically put into or clod and, and interrogate the data. And the sheet contains a row for each of the articles that we publish, including one cell, which is the full text of the article.
00:07:46:13 - 00:08:23:00
Kevin Delaney
And so it's incredibly useful for saying, like, we're just in the process of expanding our women's sports coverage as the Bay area, like, gets new teams, professional teams, the best teams in the country, the Valkyries and others. And, and one thing I did is I went in and asked, like, how should we think about the types of stories that resonate with our readers based on our historical coverage of women's sports and sports more broadly, in the context of in the context of conversions, basically subscriptions, which is like our primary sort of business goal.
00:08:23:01 - 00:08:50:08
Kevin Delaney
And the answers are really useful. And I just have never had that kind of experience with analytics before. Like, we in one way to think about it is like the readers are talking to us, like constantly in terms of what they're subscribing to read and things like that. And we just haven't had a very the ways that we've had access to that have been to like go to if your news organization has an analytics person, like go to them and say, ask some of these questions.
00:08:50:08 - 00:09:14:14
Kevin Delaney
And like three weeks later, they're generally like beleaguered. They might get back to you. They might have misinterpreted your question in any sort of back and forth around these questions, actually is very clumsy. So that that actually is probably the most exciting thing that I personally have been doing. And then the fourth area is with the support of Lean Fest and OpenAI and Microsoft.
00:09:14:15 - 00:09:54:08
Kevin Delaney
We about 4 or 5 months ago launched an AI native first app for the standard. And that's just been really exciting to have that opportunity to to use AI for personalization. We're starting to use it for audio. We're using it for synthesis, which is basically like helping readers like different sort of formats, taking our articles and translating them into formats and on and on, actually going in and finding like location references in articles and then plotting the articles on the map.
00:09:54:09 - 00:10:16:15
Kevin Delaney
Like these things. You could a lot of them you could actually do before. But the injection of AI makes it possible for a small newsroom like ours to actually start launching these features. Yeah, makes it possible key right there. And Kristen well, we're doing none of that. No, I'm just kidding. I would like to that listening to some of what you're talking about is really exciting.
00:10:16:15 - 00:10:40:14
Kristen Muller
I think we are. We're a new newsroom. We formally launched in January and we are hyper local, community centered and participatory, meaning we work not just with professional reporters, but people in the community who are sometimes paid to go to local government meetings, sometimes just show up to be part of it, part of it. And we also work with about 50 high school journalists across the city.
00:10:40:15 - 00:11:10:02
Kristen Muller
So we're all working together. And our mantra is service over status. So everything we do has to be in service of a specific resident need in the neighborhoods that we're at. And that is our theory as to how we build trust in terms of providing news and information that can help Angelenos improve their communities. But that trust is super fragile because of the places and populations we're serving.
00:11:10:02 - 00:11:48:04
Kristen Muller
So we are serving mostly immigrant communities on the East side. And in downtown, just south west of downtown, in Koreatown, Pico Union, Westlake, and then in South LA and Inglewood, which are also communities which have been traditionally underserved by mainstream media here. So we are really we, of course, use it for things like transcriptions, although, you know, even that has been something we've questioned because we don't disclose to the people that we're interviewing ahead of time, that we will be putting their remarks into Otter and uploading them to a cloud with their names.
00:11:48:04 - 00:12:29:04
Kristen Muller
And a lot of the people we interview are undocumented. So we've been starting to have conversations about how do we make that use transparent. We've also used it for translations, which is really, again, when we think of service, it's how do we for breaking news, especially if we're short staffed and we are trying to get the information out in multiple languages we've used it for for translation, most specifically in Boyle Heights, started with the fires back in 2025 when we were trying to get information out quickly about air filters and other resources that community members had access to.
00:12:29:04 - 00:12:56:10
Kristen Muller
And since then, we we've continually used the translation service that AJP helped us develop or actually helped another newsroom develop and then connected us with them. So we see a lot more potential in that because we are serving we are based in Koreatown, which obviously, you know, is there are many people there who are Korean first in terms of language and also Filipino town.
00:12:56:10 - 00:13:25:08
Kristen Muller
So Tagalog is another area we're looking at. I think we've actually just as a newsroom when you ask like, what do you what what's the impact it's having on newsroom? I would say like pulling our hair out because we are trying to understand as fast as we're experimenting and it is nearly impossible. I mean, literally, you know, every morning about if we opened up the New York Times app right now, there be an AI story, you know, in the first before you had to scroll.
00:13:25:08 - 00:13:34:04
Kristen Muller
So it's really evolving quite quickly. And one of the values we're trying to.
00:13:34:06 - 00:14:00:07
Kristen Muller
Inculcate as a culture is learning. And like how are we learning from what we're doing? How are we? What did we just do that we can learn from? And so when the tools are outpacing our ability to learn, it gives me a little pause because I don't feel confident going to a community member and saying, you know, your data is safe with us when I can't trace that data across different places.
00:14:00:07 - 00:14:37:11
Kristen Muller
So I think that is where we are trying to upskill really quickly. And luckily we've had a lot of support in that area. The other use case that we tried, which again prompted a lot of conversation with us, but we did it quite quickly. So this was during the lineage fire down in East LA. And our team, Boyle Heights, had been covering it around the clock, and one of the things we had done is asked community members what questions they had for Mayor Base, and we didn't know if we were going to get to put the questions to Mayor Base.
00:14:37:11 - 00:14:57:20
Kristen Muller
We just thought we should keep a list running so that when we get the chance, we can whip it out and just ask. So we had put out we we had actually put out like shoeboxes in the recovery centers, which I joked they were wrapped in newspaper and looked a little Unabomber, but like so people put questions in them, which I was surprised by.
00:14:57:22 - 00:15:17:22
Kristen Muller
And then we had Instagram posts, Facebook posts saying, if you had a question for the mayor and in our newsletter. So we were soliciting all these, these questions, and we got about a two hour heads up that the mayor would talk to us. So, you know, we had all these questions and everyone's out in the field doing their things.
00:15:17:22 - 00:15:46:07
Kristen Muller
And so we, you know, made it called an audible and put them all into ChatGPT and said, you know, identify the themes and then suggest some questions. And we went over them as a group and tweaked the questions, but we would not have been able to synthesize that information within the two hour window we had. So that was a case where I felt as I made the call, I was like, we're just going to do this.
00:15:46:09 - 00:16:08:01
Kristen Muller
And I felt I could justify it in the moment because the questions didn't have names, there wasn't any personal information, and we were able to kind of close the loop with the audience. Your questions asked. So that was the other time we've used it in a way that we all felt pretty good about, I think. But we're in the process of developing our policy right now.
00:16:08:01 - 00:16:34:21
Kristen Muller
And, you know, we have the luxury of not we're not shrinking as a newsroom. So we're not our staff doesn't come at it from a point of whose job is going to be eliminated first, which is a tremendous privilege to be able to have right now. So I think we're able to have a kind of thoughtful, collaborative, consensus driven process, which is also new, a little bit new for me in terms of a newsroom culture.
00:16:34:21 - 00:17:00:09
Olivia
So that's where we're at. Thank you. So I want to go back to the pattern. I started with how every time technology made one part of the job easier, the newsroom didn't give the reporter the time back. It gave them another job, which we've hit on a little bit. But a specific case in point a couple of weeks ago, an anchor at Fox 26 Houston posted a clip of herself anchoring the station's live stream while doing the entire control room by herself.
00:17:00:15 - 00:17:26:06
Olivia
Has ever been here? Seen that clip? A lot of us, yes. So here's the question is AI the thing that finally helps her, that switches the cameras and rolls the graphics so she just has to anchor? Or does it keep going until it takes the anchors job two and replaces her help or harm? Kevin. You first. I'm sorry, I don't feel expert on the on the TV use case.
00:17:26:08 - 00:17:51:08
Kevin
What? The scenario that seems most likely to me is they're like in a broad brush. They're like two types of journalism that we do. One is I think Gideon Litchfield it talks about is like summarize. And so these are news that actually can be very well represented in summary like that you could in a in an AI tool.
00:17:51:08 - 00:18:20:15
Kevin
And I feel like those that more like commodity generic summarize news is actually quite vulnerable to people just deciding to get the information from Google search that has an AI summary or from ChatGPT or Claude. And so there's a there's a lot of journalism like that. I did a lot of that when I was growing up as a reporter, and that stuff just is not differentiated.
00:18:20:15 - 00:19:03:00
Kevin
And I'm not super sentimental about it. The part that I am concerned about that is when news organizations are investing resources in breaking news that can be summarized, and when that is summarized without any economic benefit, traffic, anything by AI tools. So there's a breakdown in those cases in the like, there's no incentive to actually break news or introduce new information into the corpus, as people say in Silicon Valley, if, if, if the value is actually captured by the AI companies.
00:19:03:00 - 00:19:43:13
Kevin
So that's one concern I have there, the part of news that I actually feel quite optimistic about is the other part of it, which is things that are investigative, narrative, opinion, ideas, analytical, the kind of the most fun, interesting journalism journalism of like proximity. I think like one analogy, which is probably deeply flawed, but is like if you think about food, you have like the industrial farming, like is is the summarize information and then you have the like organic farmers market, food co-op stuff, which is higher nutritional value.
00:19:43:14 - 00:20:06:07
Kevin
It's based in proximity, it's premium priced, it's more interesting and quirky. It's like a more interesting tomato than you get at your local like, mass supermarket chain. And so you know that, like among the things I'm trying to do at the San Francisco Standard is to try and just make sure that we keep focusing on original journalism. That brings perspective.
00:20:06:08 - 00:20:38:15
Kevin
Benefits from proximity is not easily reduced to three bullet points in someone else's app or interface, basically, and I actually feel very readers I see from the subscriptions subscription dynamics is it? That's actually what people subscribe for. That's what people want. There is a market for that. Neal may very much agree some of these. I was at an event full of publishers, and one of the publishers up in the front row was complaining about Google Zero, right?
00:20:38:16 - 00:21:02:17
Neil
The fact that AI searches are taken away all the traffic, and it was somebody from a magazine and the entertainment business that does a lot of garbage content anybody could find in a web search. And I kind of I was sitting there thinking, if you built your entire business based on the business model idea that Google will never change, you're probably still smarting from your Facebook relationship, right?
00:21:02:18 - 00:21:22:06
Neil
And I don't know what before that, it just, you know, it's incumbent on us to do the kinds of stuff as journalists that we can do that nobody else can do. And if you're still producing, what time is the Super Bowl like? I don't think you're really doing the thing that we're here for. One of my investigative reporters is here, right?
00:21:22:07 - 00:21:51:19
Neil
That process of banging on a bunch of doors to try to find a story that you find out doesn't exist, but finding the other one that's really interesting and spending months and months doing that, the tools are going to help a lot. And they do. We, you know, the idea of replacing work, the more, the more a long term, project based investigative your work is, the more I think this tool is going to help you a lot, but you still have to be there and drive it.
00:21:51:20 - 00:22:10:00
Neil
If a lot of the routine stuff becomes easier to do, if we can produce, you know, with this digital democracy database, we set out thinking this is going to be a product that's going to do all these deep campaign finance stories. It's going to find an amazing story for you that you're then going to pursue and write about how money is affecting politics.
00:22:10:00 - 00:22:25:20
Neil
Some of that is happening. We're now giving it to journalists all over the state. Most of what they're using it for is a much simpler story. There's a bill that's really important to my town that I didn't know about until I got this amount of data from the Digital Democracy database, and now I can write a short story about that.
00:22:25:20 - 00:22:41:23
Neil
And sure, why don't we maybe write the short story and give it to you because you don't have time to do that anymore. There's a lot that can come out of that. We're realizing this thing is not just a product, it's a platform that can do all kinds of things. But the if we do this right, we have the ability to improve the quality of the work we're doing.
00:22:42:00 - 00:22:59:19
Neil
Stop doing the routine. My first job at the San Francisco Examiner as a copy editor, I would edit a list every day of all the ships coming into port in San Francisco that day. It came in by fax, which was a fancy new technology at the time. Like, if a journalist doesn't have to do that, we don't have a whole lot of extra people right now in this business.
00:22:59:21 - 00:23:23:12
Neil
If that journalist in Houston doesn't have to do all the switching and everything and can just think about local journalism, why can't we build her entire show with tools and put that on the air? Right. The, the and I don't mean to suggest all the technician jobs should go away, but the more that person is out in the community knocking on doors, talking to people, learning things that no tool would have known to look for because we didn't know they existed.
00:23:23:14 - 00:23:48:13
Kristen Muller
That's that's the value, right? Kristen. Yeah, I mean, I agree completely with both of them. Said, but I'm still thinking about that poor Fox producer. How stressed she must be. I also think like how you phrased it. Is it? How did you phrase it is a win or a loss? Would it help or harm? Help or harm? It's like to who you know and like, does it help her or harm her?
00:23:48:13 - 00:24:12:15
Kristen Muller
I don't know, but like big picture wise, I think you're making a really effective case. Like there are things we know that it's absolutely going to help us do. What I worry about is the gap between what we know now and what we don't know. And that to me is where, like, yes, it's very easy to help with data sorting and scraping and organizing and yes, thank God.
00:24:12:15 - 00:24:34:10
Kristen Muller
I mean, I remember the days we had a librarian, you know, and you used to call her at nine in the morning to warn her you were going to have a question coming in at like 1:00, that you were going to need the answer by three. And obviously we don't do that anymore. But I, I also just don't I, you know, I think like how do we, how do we set up like checks and gates.
00:24:34:11 - 00:24:51:15
Kristen Muller
Like what are those filters we need along the way as an industry, not just newsroom by newsroom. And that's a conversation I think we're really beginning to have an earnest, especially like venues like this forms like this. But there just has to be so much more of it, and there has to be so much more of it, like tomorrow.
00:24:51:15 - 00:25:16:03
Kristen Muller
So, you know, I'm really lucky I can call Neil or text Neil or now I'm going to call you and say, what's your what's your opinion on this? And I get really smart takes quickly. But like for an editor that's maybe working in a rural part of the country and who doesn't have that kind of network and is making these choices on their own in areas where there is so, so much less information?
00:25:16:05 - 00:25:40:17
Kristen Muller
That's where I begin to worry. So like, how are we distributing the knowledge and having consensus, not just with in journalism itself, but like, again, in participating in participation with the communities we're trying to serve. And so that's a framework I would like to engage in. And I think like that's a whole that you're trying to fill. So thank you, Neal.
00:25:40:22 - 00:26:04:21
Neil
You've led newsrooms as an editor and now run Cal Matters as CEO. Where are you finding the most useful applications of AI on the business and operations side, and how do you decide which uses are worth investing in? That last part is fascinating, right? Because you don't necessarily know what to invest in until you've tried it out and you got to you have to make an investment just to to try it out and do it.
00:26:04:23 - 00:26:26:13
Neil
We are we are certainly using the tools. We have a fundraising team. Their job is to identify the people who are most likely to support the work we do, and go find those people and knock on their doors over and over and over till they finally mistakenly open the door once. And then you can talk to them and the research you're trying to do about who the prospective donors are, where they've given money before.
00:26:26:14 - 00:26:42:06
Neil
All that kind of stuff is certainly made easier by these tools. We're also able to use the tools to do a better job of identifying you when you come to our website. If you've chosen to let us know who you are so that we can personalize it and not put in information that we're not going to put a newsletter subscription up there.
00:26:42:06 - 00:27:03:05
Neil
If you're one of the quarter million people already gets our newsletter right, we're we're trying to make this stuff more useful. We are building a newsletter out of this digital democracy database called My Legislator. You get it once a week, and it tells you what your two members of the California State Assembly and House Assembly and Senate actually did last week.
00:27:03:07 - 00:27:26:08
Neil
And is that a business case or an editorial case? It's a newsletter. It's a product. Newsletters encourage people to interact with us more. They leave more people to become members. So it's an editorial product. It's also a business thing. And in many cases those are now combined, right. We're doing it for both reasons. That newsletter is fascinating because it has proven to us we we invested a bunch of time and effort into it.
00:27:26:10 - 00:27:48:17
Neil
Somebody built a very handmade, clunky way to do it, and I won't call him out, but he's in the back row and made this thing happen with, you know, we weren't sure if we could do it. And like, he sat up all night hacking at it and made it work. We're going to clean that up sometime. But it it has proven something to us that's really valuable, which is yeah, people want to get newsletters and this newsletter will be will be useful.
00:27:48:19 - 00:28:04:04
Neil
This newsletter tells you how your state legislator, the person you voted for, the person you're paying their salary, you sent them to go work for you in the state capitol. They're your employee. Basically, you and the other half million people in your district. You're not paying any attention to what they do. You might care, but you have no time.
00:28:04:04 - 00:28:19:10
Neil
You're not going to go look it up. When we do a newsletter that drops it in your lap every week on Monday morning, you look at it for a few minutes and people are reacting to it. They're writing letters, emails to their state legislator and copying us and say, hey, why did you vote that way on that bill?
00:28:19:11 - 00:28:48:15
Neil
Or why didn't you vote last week? What the hell were you? And that is proving to us that there is a place for using these tools to give people information they would be willing to engage with, to become more engaged in their democracy. If you just made it easy for them. And to us, that is as much a business thing as it is editorial, our, our membership and all that newsletter stuff and everything is in the editorial side of the operation because it's all about serving the readers, which then leads to membership.
00:28:48:17 - 00:29:06:17
Kevin
Kevin, we know that you've co-founded Courts and Charter and now lead the San Francisco Standard, drawing on that experience, what makes an AI tool or app useful enough for journalists to adopt in their daily work? And how is that shaping your experimentation at the standard now?
00:29:06:19 - 00:29:30:07
Kevin
It's a really good example, a really good question in the context of tartare. We interviewed some executive who's talking about how AI tools can't be a swivel chair, like you can't be doing your work and then like, have to use this other thing. It just has to be kind of in the flow of your work. And so. I think that's what we're honestly, we're just trying to figure that out.
00:29:30:08 - 00:30:05:09
Kevin
And the most research shows that the most important factor, one of the most important factors in AI adoption is, is managers, leaders of organizations actually modeling that they use AI tools and like showing not just saying that they use them, but actually showing people how they're using them. And if you don't do that, people either don't use the tools or they actually hide them because they feel like they're people are judging them for cheating or like that kind of thing.
00:30:05:10 - 00:30:27:16
Kevin
And so I guess the thing that I've tried the most to do is like with some of the analytics and other things that I was talking about, just to like in, in the context of the newsroom, talk about how I'm using them, talk about what they're good at and what they're not, use them like sort of integrate them in the flow of actually like how the newsroom operates.
00:30:27:16 - 00:30:53:21
Kevin
And I find that reporters will come up to me like, hey, I you know, I did this thing and like using AI. And that for me is like a signal that that it's working like you took away the stigma, right? Yeah. Kristen, the LA local has grown from its roots in Boyle Heights to serve communities across LA. Where could AI help a small reporting team cover neighborhoods or stories?
00:30:53:21 - 00:31:01:10
Unknown
It doesn't currently have the capacity to reach?
00:31:01:12 - 00:31:30:11
Kristen Muller
I don't know, we think this is like a very active conversation, so we know that there are, you know, like hundreds of local government meetings happening all the time across these areas that we cover. We have a program called documentaries, which actually trains and pays residents to go to those meetings and take notes. Now we have just, I think, just under 300 of them at this point.
00:31:30:13 - 00:32:06:23
Kristen Muller
So, you know, we're got a pretty robust army of civic stenographers at this point. But there but they are, of course, you know, only so many of them, only so many meetings. And we've had a bunch of companies pitch us on, hey, we can scrape these local meetings for you and turn them into briefs. We have not done that yet, but I think that's not to replace the documentaries because honestly, some of the stories that come out of our documentaries coverage have nothing to do with what happened in the meeting itself.
00:32:06:23 - 00:32:30:12
Kristen Muller
But like the chat outside the meeting. So they're not they're not going to be replaced in that sense. But I think it might give us more, more breath. I don't know if it's depth necessarily, but it might give us more breadth. Of course, it also requires an editor or human to review the transcript with all the other stuff.
00:32:30:14 - 00:32:46:07
Kristen Muller
And to be honest, I think it's nice to have that information, but kind of to what you're saying, Neal, people are busy. No one wakes up wondering what happened at the zoning committee last Tuesday. And except for the three, unless they approved a giant building next door to my house. Well, yes, I really want to know. Right? Yes, exactly.
00:32:46:07 - 00:33:13:20
Kristen Muller
So there's that human element of it, right? Like, how are you going to make the connections between what just happened and who it actually affects? But that's an area that I think, you know, we're actively talking about. And the other is just kind of kind of the standard workflow that you're talking about. You know, we're all on slack all day, and we have a lot of different neighborhoods we cover and a lot of people we talk to.
00:33:13:22 - 00:33:37:14
Kristen Muller
And right now, a lot of that knowledge lives in the heads of each individual employee. And whenever we get together and say we share notes from the community listening, we're doing and it's like, oh, you talked to him. I talked to him three weeks ago. And, you know, so like maybe connecting some of the dots and finding more themes across what we're hearing from people in a structured way is, is another area I'm thinking about.
00:33:37:15 - 00:34:13:13
Kristen Muller
And I mind you, I don't think any of my staff is in this room, which is probably good because I haven't actually articulated this allowed to anyone. You're the first. But we do have, you know, we just have a lot of information from the community coming to us in a way that is not structured right now. And so figuring out ways that we could, again, not using people's names, you know, age locations per se, but finding a way to group them so that where our reporting is more tailored to individual concerns, I think would be something we're very interested in real quick.
00:34:13:14 - 00:34:29:03
Neil
One of the privileges of being the statewide nonprofit newsroom, right, is we're we're thinking about those same issues statewide. And if we play this right, as we've done with the LA local and other friends around the state all the time, we should be able to build that solution once for local meetings for California and probably the whole country.
00:34:29:04 - 00:34:49:21
Neil
Kristen is the person who overnight the first night of the Altadena and Palisades fires, who, by a series of text messages, created a daily newsletter that went to 900,000 people every day about the fires for three weeks just by sending out some text messages to all of us saying, let's do this. And there wasn't even any AI involved in that, right?
00:34:49:22 - 00:35:10:16
Neil
If you take the the instinct to work together and figure out each other's strengths and build stuff together, then you throw in the AI that unfortunately it's not a breaking news story, so we won't get it done in a day like we would if we're under pressure. We do much better work under pressure, but being able to build those things together, I think, is part of the magic here that makes it possible, especially with everybody having smaller newsrooms.
00:35:10:18 - 00:35:16:22
+ Read More
Like
Comments (0)
Popular
Watch More
AI in Newsrooms: Responsible Innovation and the Future of Journalism
Posted Mar 23, 2026 | Views 907
# Journalism
# AI Adoption
Thinking Machines & AI Economics: How Reasoning AI Is Rewriting the Future of Work, Science, and Strategy
Posted Apr 23, 2025 | Views 3.6K
# AI Economics
# AI Policy
# AI Research
# Future of Work
Event Replay: Using AI to Protect Children Online: In Conversation with Thorn
Posted Feb 11, 2026 | Views 1.8K
# AI Safety
# Ethical AI
# OpenAI Presentation