Selling AI to Government: Pretend Like It’s the Early Cloud Era (c. 2010s)
Government agencies are sitting on billions in federal funding earmarked for modernization, but most of them aren't buying AI – at least not yet.
We spoke with senior sales and strategic leaders at three companies selling AI software and implementation services to the public sector – two are smaller startups, and one works is a global business and technology consulting firm with deep experience implementing cloud services.
Here's the reality, and what it means for anyone selling technology to the public sector.
Government Isn't Buying AI. It's Barely Using What It Has.
The current state of AI adoption in government is, to put it directly, extremely limited. Most agencies have access to tools like Microsoft Copilot through existing enterprise agreements, but they're using them incorrectly – applying them to tasks the tools weren't designed for, with no governance framework and no direction from leadership.
The result is a shadow AI problem: individual employees experimenting on their own, often inappropriately, while agency leaders remain disengaged. The gap isn't technical access. It's that most agencies don't understand what problems AI can actually solve beyond chatbots.
The Reframe That Works: Call It Automation
One of the most effective tactics for breaking through resistance is to stop saying "AI" altogether. The word triggers skepticism, political caution, and bureaucratic inertia in government settings. Framing the same capabilities as "automation" – and leading with specific, tangible use cases like student attendance tracking or early intervention workflows – lands differently.
This mirrors a pattern we've seen before. Cloud adoption in government followed a nearly identical resistance curve more than fifteen years ago. The private sector moved first because cost savings and efficiency arguments were enough to drive decisions. Government, by nature risk-averse and consensus-driven, lagged by years. AI is tracking the same trajectory, with the same inevitability.
The Sales Timeline Is Real: Plan for 12 to 18 Months
Selling to state and large local governments is not a quarterly game. The typical cycle runs twelve to eighteen months end to end: roughly six months spent convincing an agency that they have a problem worth solving, followed by another eight months navigating the path from decision to signed contract.
Budget cycles matter enormously. June and July are peak procurement season for many states, which means right now – late winter and early spring – is the ideal window for budget planning conversations. If you're not in the room when agencies are mapping next year's spend, you're already behind.
Who You’re Selling to Matters: CIOs vs Administrators and Agency Heads
The approach that works with a state CIO will fall flat with a business leader in the same department or agency, and vice versa. CIOs typically know exactly what they want and which platform they prefer. The conversation is technical, specific, and often competitive.
Administrators (e.g., a city manager) or the head of an agency on the other hand, know the outcome they're trying to achieve but need help mapping the process and technology path to get there. They want a partner, not a pitch.
The best sellers are technology-agnostic, leading with problem identification rather than product capabilities.
The Bottom Line
Government AI adoption is coming, but the sellers who win won't be the ones pushing the technology. They'll be the ones who understand the agency's real problems, speak the buyer's language, respect the timeline, and make it easy to buy.
The playbook isn't new – it's the same trust-based, access-driven approach that has always worked in public sector sales, applied to a category that government is just beginning to understand.