← The Daily Scoop Podcast
The Daily Scoop Podcast
The future of AI in the federal workforce
00:20:35
If you ask Office of Personnel Management Directo…
Loading summary
Transcript20 lines
- [00:00]
A
Today on the Daily Scoop Podcast from the Scoop News Group, the future of AI in the Federal workforce. Director Scott Cooper shares the plan for AI from both Big OPM and Little OPM vantages. It's Tuesday, May 12, 2026. Welcome to the Daily Scoop Podcast where you'll hear the latest news and trends facing government leaders. I'm the host of the Daily Scoop Podcast, Billy Mitchell. Thanks so much for joining me. And now let's dive into the day's top headlines. The Department of Homeland Security intends to continue its work with cellebrite, a provider of digital forensics hardware and software tools. According to forecast documents released last week, Immigration and Customs Enforcement, as well as the Department's Homeland Security Investigations Unit plan to award a five year indefinite delivery indefinite quantity contract with $100 million ceiling to to the vendor later this year. Cellebrite's products enable the agency to access data from cell phones, tablets and more recently, unmanned aerial vehicles, or UAVs. The Israeli firm's data extraction capabilities are the most widely utilized and deployed computer forensic tool within hsi. Per the document, cellebrite has been deployed across dhs, including its reported use within the Secret Service to break into the phone of the man who shot President Donald Trump at a campaign rally in Butler, Pennsylvania in 2024. The vendor provides its data extraction tools to around 7,000 customers across the public and private sectors, including federal agencies outside of dhs. The Department of Justice recently served as the sponsoring agency for Cellebrite's pursuit of a FedRAMP high authorization, which it achieved last week after an 18 month process. Now, in other news, Doge's playbook for using artificial intelligence to eliminate regulations was on full display at the Department of Housing and Urban Development last summer with the introduction of an AI tool built for the extermination of federal housing rules. Documents obtained by Democracy Forward via Freedom of Information act requests reveal a PowerPoint presentation delivered at HUD on Sweet Rex, a tool named for DOGE associate Christopher Sweet, according to Wired reporting last August, the new documents shared with FedScoop laid out a multi step process in which all HUD regulations would be analyzed by the AI. The tool would then provide recommendations to keep delete or partial delete each rule per the presentation. Attorneys would review the suggestions and agency staffers would make the final decision. HUD regulations cover everything from the prohibition of discrimination on the basis of sex and mortgage assistance to providing legal aid for foreclosure related issues. Nevertheless, a source familiar with the FOIA submission said in an interview with FedScoop that the explicit instructions given by DOGE to HUD employees make clear that the default assumption is is that rules and regulations should be rescinded, the source said. It's not clear from the documents if any HUD rules had been written or rescinded based on SuiteRec's recommendations, nor is it something that can be determined by looking through the Federal Register. But the general feeling among those familiar with the FOIA submissions is that LLMs are not equipped to perform this type of analysis and outsourcing this work to an AI tool undercuts the role that Congress has empowered agencies and humans to to play. HUD's press office did not respond to FedScoop's questions about the tool, including whether it was deployed at the agency and the extent to which DOGE is involved in the agency's operations. Doge, meanwhile, made the case in its presentation that HUD's attorneys and policy teams would be freed up to do more high value work thanks to the AI tool, which claimed to slash rulemaking time by at least 90%. The pitch to HUD employees from DOGE was that the tool would simply automate all the most time consuming steps in deregulation and leave program groups in control every step of the way, along with reviews and edits from legal teams as needed. For more news at the intersection of the federal government and technology, make sure to visit fedscoop.com if you ask Office of Personnel Management Director Scott Cooper how he thinks about the role of AI in his agency's mission, he'll tell you he sees two different prevailing perspectives, one with a big OPM mission and and another for little opm. At least that's how he described it to me recently at UiPath's Fusion Conference in Washington, DC. During our onstage interview, Cooper shared that juxtaposition, as well as some emerging use cases for AI that OPM is driving forward and much more about the future of the federal workforce and how AI will play a role in it. And now let's go to that interview with OPM Director Scott Cooper.
- [04:50]
B
Scott, welcome to the stage.
- [04:51]
C
Thank you. Glad to be here.
- [04:53]
B
It's like we were doing this just yesterday.
- [04:55]
C
I feel like I've seen this room before. It's good. Hopefully we have a different audience this time.
- [04:58]
B
We do, I think, and it's exciting that we get to dig into, you know, a lot of what we talked about before. But I heard there was some news that came out of OPM yesterday in terms of AI announcement of sorts. I think we'll get to that a bit, But I'd like to start because we've talked previously, and one of the things that stood out to me in our former conversations is the way you've sort of juxtaposed the mission of OPM in the context of little OPM and big opm. And I know that OPM has a massive mission in either sense, but I'd be interested to hear the way you're handling and navigating those two dynamics and particularly how AI can play into that.
- [05:34]
C
Yeah. First of all, again, thanks for having me, Billy, and thanks for everybody here. We'll try to keep this on time so we don't keep you from lunch, obviously. So, look, here's how I think about it. So big OPM in my mind, for those of you guys who don't know, is we essentially set healthcare, excuse me, human policy, talent policy for the federal government. And then we're also kind of what I would consider the group that makes sure that we have the right seats in the right, you know, kind of places from a talent perspective. So we're trying to then model that behavior inside of opm. So I think the things that we've got to do with respect to AI, number one, is a talent issue, as you and I have talked about. So, you know, we've got a number of different engineering roles within government, but there's no question that there are modern software development, modern AI roles, data scientists. There's places like that that we're missing talent. And so one of the big things that we need to do, both within OPM and broadly across the government, is figure out how do we address that talent gap. And we can certainly talk about that. And then two is we've got to make sure that people view AI as an enabling technology to sit alongside talent and help them actually augment what they're doing. So I think too much of the conversation today is always about, okay, what is the role of AI and what will happen in jobs in the future. From where I sit, look, there is an endless demand on people's time, whether it's in the federal government or whether in the private sector. And so I think people ought to rightly think about AI as if there are ways in which you can find small ways to improve efficiency in your job. It frees up time to do the things, quite frankly, that we never have time to get to. And so the more we can actually have people doing the things that they are uniquely qualified to do, and quite frankly, have AI help improve the efficiency there, I think we're way better off in terms of delivery that we can do on behalf of the American people.
- [07:06]
B
That makes total sense, and I think we can talk a bit more about that. But before we do, I do want to go back to the news of yesterday. Usa, USA Class is it called? And it's, it's a new application of AI for the hiring process and job descriptions for the federal government. But I'm sure you can explain it a lot better than me, so why don't you share a little bit about that?
- [07:25]
C
Yeah, so what we announced yesterday is, in my mind at least, a very good example of what we're trying to do broadly with AI. So in 30 seconds or less, when you're hiring people in the federal government, it's a very complicated process of writing the job descriptions, making sure those job descriptions deal with what are called the classification standards, which is, what level should that job be, what job family does it belong to? And look, we've just got a lot of jobs in the government, right? We have 600 plus classifications. We've got 2 million employees. And so, like, the N factorial of this is pretty significant. So what we've done with AI, I think is actually very simple, which is we basically said, okay, what is AI good at? Well, it turns out if you feed AI a bunch of job descriptions and we put a prompt in front of it, it's a lot easier to create the next job description by you being able to type in to say, okay, look, I'm hiring for a financial analyst. Tell me about the skills that are needed. Right? It's all very plain English stuff. And then we can kind of spit out at least a draft of a position description. Now again, I know this sounds obvious, but to me, this is exactly what we should be doing in AI, which is we're not replacing humans here. We're not just having AI create like magical job description. But what we're saying is if the AI can actually learn from what we've done in the past and it can streamline the process, I'd much rather have my HR hiring managers focused on outbound recruitment or talking to candidates or stuff like that, rather than having them mired in a bunch of paperwork around a description. So to me, again, this is like, you know, we're not, we're not, you know, you know, creating rocket science here or, you know, curing cancer basically, but we're just figuring out ways to be smart about how do we use people's time and where computers, quite frankly, are good at reading through and, you know, kind of parsing Large volumes of data and helping us collate stuff in my mind, like we should be doing this all day long. So this is just kind of, you know, iteration number one of what I think we can do to improve the efficiency of the hiring process.
- [09:05]
B
Absolutely. I've been following the space for more than a decade. It's something that folks have been talking about for that amount of time or longer. And it's great to see it finally getting done because it is the low hanging fruit there. We talked a little bit about the larger context of the transformation of the federal workforce and how AI and the federal workforce is dealing with the incoming AI, if you will. Obviously folks are going to be a bit worried about that, as you've expressed. But how do you convince folks that this is a good thing and you know, particularly those that are maybe laggards or hesitant and don't want to get hands on with it, how do you give them opportunities to get familiar with it so they see the utility and the optimal uses of that technology?
- [09:46]
C
Yeah. So look, I think there's a couple things we have to do. One is just from a cultural perspective inside of government, you know, this having been around government for a long time, is we've had this general philosophy of, you know, we take no risk. You know, people are very worried about doing something wrong. And we also, quite frankly on the other side, don't often give people kudos and or promotional opportunities around upside opportunities. So part of this is just a cultural reset to say, look, when you're using new technologies, there are going to be things that we make mistakes for. And the goal can't be to make zero mistakes. The goal is to be able to identify what do we think the risks are. And let's have that conversation. And yes, if we are working at Department of War and we're talking about the risks of people dying or not being able to defend our country, then maybe the answer is yes, it is zero risk in that area. But the reality is certainly in my organization, a lot of what the government does, the answer in many cases is if we can balance a reasonable set of risks with significant upside that we think comes from using new technology, we have to just culturally create that permission for folks. So I think that's thing number one is just making sure people understand that anytime you're dealing with new stuff, we have to give people permission to experiment in a way that again gives them that freedom. Two is, look, I think the most important thing is giving people also freedom from a time perspective to experiment. So this is a Tiny thing. We just, Politico just put a story out based on an interview I did. We have literally five LLMs on people's desktops, which is probably crazy and makes no sense because most people probably aren't going to use five LLMs. But we did that mostly just again as a cultural statement, which is, look, here's a bunch of alternatives. Try one, see what you like. Again, I'm not asking you automate your job away. I'm asking you to say what is 5% or 10% of what you do today where you could benefit from being able to type into an LLM and have it do some basic research for you or, or have it summarize documents for you, Stuff like that. Right. So I think what you need to do to demystify the worries is incrementalism. And again, I know that sounds kind of weird to say, but like we don't need a 2040. What does the 2040 OPM business plan look like for AI? But what we need is like what are you doing today to figure out ways in which you can actually improve efficiency for a couple percentage points at a time. And then I think people get familiar with it and then what we will see coming out of that is yes, okay, now there are full on applications we should build to incorporate into that. So you know, we run the retirement services operation and that's been a horribly manual process for a long time. And so yes, like we can incrementally add AI to help us be smarter about it, to help us improve efficiency, to help improve quality. Like that's the kind of incrementalism I think we need in the government. But it starts with giving people permission to recognize that, you know, not everything is going to be 100% right. And there are some areas where it's okay to make those trade offs as long as we do that. Eyes wide open.
- [12:23]
B
Yeah, it starts with a small first step to get on that journey. You mentioned retirement services. I know that's a area that can be a pain point for folks as they're leaving the federal government, but one that is ripe for modernization and automation in the sense of AI. So tell us how you and your OPM are working on applying AI for customer service around retirement services processing and what people in the federal government can expect out of that.
- [12:49]
C
Yeah. So retirement sources, for those of you who don't follow it basically has been I think the bane of every annuitant experience in the federal government going back probably close to 50 years. Now the quick summary is we literally have A complete manual paper based process where we send documents by snail mail from HR departments and payroll departments and others to ultimately show up into an underground mine that we have in Boyers, Pennsylvania. So this is like, it actually would make a great like, I don't know, sci fi story or something like that. At some point in time, you know, there'll be something fun to say. So there's just basic obvious stuff which is, look, we know that people being able to use online forms to actually do an application process, a way more efficient process. And that's not. There's no AI involved in that. Right. That's pretty simple. The second piece of it though is we also have a call center obviously where we're dealing with all the annuities post retirement. They have to deal with changes to their life cycle or somebody gets divorced or there's a death in the family. And again, literally, you couldn't have designed a call center that is more living in kind of the old world. We have literally every single thing that people need. They have to pick up the phone and call. And not surprisingly, we have way more calls than we have people who can actually service it. And so the answer to this is yes. Like there are ways in which we should introduce AI. Now we have a particular population obviously of, you know, kind of retirees. So you know, not every 90 year old is going to be able to kind of, you know, talk with an AI bot. But the kind of technical fluency I find of people is way greater than most people give them credit for. And there's no question if we could actually give you a very simple AI bot that you could communicate with via chat or ways in which we can automate the call center operations so that you don't have to talk to a live agent, but you can actually do the things you need to do more. Basically, it's just obvious that that's a better solution. And again, none of this is like we need to replace people. It's literally about making sure the people we have are there for the things that only the people can do. Right? So if you've got some complicated problem, then yes, I want you to talk to an individual. But if you're trying to change your address on a retirement form, it's just so obvious that we need to do that in a way that is automated that people can do from their pajamas at midnight online instead of actually having to wait between 9 and 5 Central Time to call somebody. So that's what we're doing. Again, as I said, none of this is Rocket science. Basically it is just a basic run of the mill application of where computers can actually help facilitate customer support in a way that we just haven't done previously.
- [15:04]
B
Well, it's fantastic to hear. I know that it's an epic pain point for a lot of folks as they leave the federal government. So kudos to you on that. Another item in the news recently, it sounds like GSA is becoming a roommate, if you will, with OPM moving in with you guys. But it strikes me as an interesting potential opportunity for some collaboration. And GSA is seen across federal government as this hub of maybe a center of excellence or a technology expertise. But curious what you might think about the opportunity to sort of rub shoulders at the water cooler with their technology expertise and sort of what you might be able to gain from some of that.
- [15:41]
C
Yeah, so where we're starting with as you mentioned is look, we are trying to actually just save taxpayer dollars, which is we're in a building that is only 50% occupied. They have a need for excess space. And so, you know, kind of we did what is the logical thing, which is, hey, why don't you guys move in with us? It'll be better for everybody else. So. And then they're in the process of updating a building which we will then move into sometime in 28. So like that all makes sense. But I agree with you. The similarities in my mind between OPM and GSA is they're the two kind of cross functional agencies within government. And you're absolutely right, which is, look, GSA focuses mostly on real estate on acquisition, but a huge portion of acquisition of course is technology related stuff. So there's no question in my mind there will be like more synergies that will come from that. But you know, as a starting point, we're just doing kind of very simple thing of making sure that we don't literally waste money on space that's sitting dormant.
- [16:29]
B
Makes sense. So, Scott, let's close with techforce. We haven't touched on it yet, but it's a massive part of this administration's push to bring new tech talent into the government. We've talked quite extensively about it in the past, but I'm curious for this audience as they think about that intersection of the private sector and bringing more of that knowledge base into the federal government. What are some of the latest developments there and what should this audience do know about the progress of that program?
- [16:55]
C
Yeah, let me just give you a quick framing of kind of the big picture stuff. What we're trying to do in recruiting. So some of you may know this, but basically the federal workforce has only about 7% of its population that's under the age of 30. And on the other side, by the way, close to half of our population of federal employees is over the age of 50. So if you just think about that demographic, if we did nothing else, we would have, you know, kind of a very large scale retirement population probably over the next 5, 10, 15 years of government. And the honest answer is we've done a really poor job of getting early career people into government. So. So we are, number one, like, trying to kind of reinvigorate the concept of early career engagement in government, trying to help people understand what we think is the compelling opportunity to go spend time in government. And techforce is one of the kind of initiatives we have there. So TechForce addresses not just early career, but also talent gaps that we have. So obviously we've been talking about AI a ton. In order to modernize and do the things we need to do. There are skill sets that we have today, you know, that we need to basically bring into government that will help us accelerate those efforts. So we need more software developers, we need more data scientists, we need more AI specialists, we need more cyber folks. And TechForce is really focused on that. The other part of TechForce though, is we're partnering with the private sector here. Because I want to change the narrative, at least that I've heard from a lot of folks in the federal government, which is in order to, I think, attract early career folks. Our pitch to them can't be come to the federal government and spend 40 years here. Now, some people may want to do that, but I think for the vast majority of people in this generation, they think in much smaller increments. And so think of Techforce really as come spend two years, three years, four years in government. Let us help you from a career development perspective. Let us introduce you both to private sector and public sector opportunities. And by the way, at the end of that time period, you decide the private sector is what you want to do. We need to embrace that and help them. And so our private sector partners will essentially be a conduit from people being in the public sector to demonstrating that the skills they develop and the work they do actually has value in the private sector. So part of this is a reorientation and really a call to early career folks to say, look, come spend some time in the federal government. Use that as a way to learn to give service to your country and develop your skill set. And then you have all kinds of options open to you whether you decide state and government to go otherwise. And so I'm really excited about it. We actually literally at OPM yesterday onboarded the very first employee on Tech first. So we're very excited about that. We've got several hundred more who are kind of, you know, have offers outstanding who are starting to map their map their start dates. So we're making good progress. But look, as always, we need to kind of continue to help people understand the opportunity of what government can do for them.
- [19:23]
B
Yeah, well, that's fantastic. Only so much we can cover in 15 minutes, but we did just recently speak on our Control Defense podcast and a lot more there. If you want to dig in some more to to that conversation about tech force, your background in Silicon Valley and other things like that, it was a great opportunity. Please give it a listen. But Scott, as always, it's a pleasure. Thank you so much for your time. Let's give him a round of applause for coming.
- [19:46]
C
Thank you, Billy. Thank you very much. Appreciate your time.
- [19:49]
A
For more on AI's impact in the future of the federal workforce, make sure to visit fedscoop.com thanks so much for tuning in to another episode of the Daily Scoop podcast, available on all podcast platforms. If you've already rated the podcast on your platform, of course, choice, thanks so much. High ratings and good reviews of the show help more people to find it. The Daily Scoop Podcast is a production of the Scoop News Group in Washington, DC. Adam Butler and Carlin Fisher help put the show together and the entire Scoop News Group team contributes. We'll be back tomorrow with more top headlines. Until then, I'm your host. As always, Billy Mitchell. Thanks so much for listening.