Charting the AI Course for the Public Sector
A few years ago, I had the opportunity to meet the CEO of a large consulting firm in Washington, DC, just as Sam Altman delivered his now-famous testimony on Capitol Hill. That event was widely regarded as the starting gun in the escalating race for AI development. At the time, conversations within consulting circles inevitably turned to AI, the new elephant in the room.
During our discussion about business goals, the CEO asked for my thoughts on Altman’s testimony and the broader outlook for AI. I shared my belief that AI’s real growth would hinge on clear regulatory lines. Once the government created a level playing field, companies could compete fairly, and Altman’s testimony seemed the beginning of that path. (A lot of us were mistaken on that timeline.) The CEO agreed but laughed when I suggested that any consulting firm inside the Beltway not already using AI would be obsolete within two to three years.
They replied, “I’m directing everyone in the company not to use it. The security impacts alone are enough to shut it down, but also, can you imagine being in government and getting a proposal written by AI? It’s ridiculous. AI in consulting will never take off.” At the time, I felt embarrassed, worried I had cost myself a professional relationship with the firm. That feeling didn’t last. Within six months, I was vindicated when a senior leader from that same firm called me to help write a white paper highlighting the benefits of AI in the Public Sector. Turns out they were preparing a proposal for the government.
I share this story not to poke fun at hesitation toward new technology but to illustrate how far we’ve come in AI’s evolution. When ChatGPT first emerged, I used it mostly to tweak emails or outline presentation talk tracks. It was far from transformative; it felt like a sharper, more conversational Google, streamlining my daily workflow.
Fast forward, and AI is indispensable in my everyday work: planning meetings, designing presentation graphics, developing app concepts, even organizing vacations. It is everywhere. If you’re not using AI, you can be sure it’s using you. AI companies harvest massive datasets, your data, from the internet to train their large language models.
Despite AI’s revolutionary promise, adoption in the public sector remains uneven. The Federal government has made notable strides, yet state and local agencies trail, with only 48% using AI tools daily, compared to 64% of federal agencies. This 16-point gap is both a challenge and an opportunity for firms positioned to bridge it. Last December, just 25% of civil servants reported using generative AI for work tasks, with one in six using such tools at least weekly. Strikingly, the usual trend in technology adoption, where federal action trickles down to states and municipalities, has not held for AI. Rather, lack of regulation has not led to wider adoption as many predicted. In a paradoxical twist, the absence of clear policy seems to discourage, rather than promote, experimentation and uptake.
Research consistently points to six key barriers to AI adoption in the public sector, each presenting a strategic opportunity for enterprising companies. Chief among them is unclear governance and ethical frameworks, cited by 35% of agencies, followed by insufficient technology infrastructure and funding at 30%. Fear of unintentionally violating unwritten rules holds agencies back.
The resource constraint is especially acute in smaller communities. Forty-one percent of survey respondents are from areas with populations under 25,000, representing an underserved and vocal market segment. These jurisdictions grapple with limited budgets, lean teams, and scant resources, making it hard to adopt new technology, but they also offer unique opportunities. Deploying new tools in large, complex urban environments is a heavy lift compared to smaller communities with streamlined infrastructure. If you want a captive and flexible test audience, start small.
This isn’t new or radical thinking, but it confirms that both my view on AI adoption and the CEO’s caution about public-sector conservatism were right. Now, with states setting their own AI guardrails, it will be fascinating to watch how far local governments will take adoption this year as they gain clarity and confidence. Early signs are encouraging as local governments begin to advance their approaches to AI governance, and given the accelerating pace of technology, even our smallest communities may soon be racing to innovate.
Many of us who have spent the better part of our careers in and around the public sector have started to pivot more significantly back to local government for this very reason. My conversation with that CEO was not an uncommon interaction in DC. In the Federal space, you tend to meet a lot of people who make a living out of quick assumptions in a risk-averse environment. We seem to spend more time talking about why things won’t work as opposed to asking, “Wow, what if it did work?” At Fractional Source we’re more interested in the promise of something working than not, and we believe small to mid-size communities offer incredible opportunity to platform new solutions. We see that as fertile ground, not just for public impact, but for real innovation.