This short piece introduces a central dilemma explored throughout the Digital Democracy Lab Handbook: the growing pressure to make democratic processes more “efficient” through technology—especially AI—and what’s lost in the process. It’s essential reading for anyone working at the intersection of democratic practice and digital innovation.

The Efficiency Imperative: Why AI Keeps Showing Up in Democratic Settings

We live in an era defined by what we might call the "optimization imperative"; a pervasive assumption that making things faster, smoother, and more efficient represents inherent progress. This mindset has migrated from commercial and industrial contexts into virtually every domain, including democratic governance, often without critical examination of whether efficiency actually serves democratic purposes.

The rise of this efficiency paradigm directly correlates with the growing influence of technologists and software engineers, not just in Silicon Valley but increasingly within governmental institutions and democratic organizations. Their optimization mindset has become a dominant framework through which we evaluate democratic systems themselves.

This creates a fundamental misalignment. Democratic processes are intentionally designed with what appear to be inefficiencies: separation of powers, checks and balances, extended deliberation periods, multiple rounds of review, protest. These aren't design flaws to be optimized away; they are essential features that prevent hasty decisions, create space for diverse perspectives, and enable the kind of slow, careful collective sense-making that effective democratic exchange requires.

Yet we increasingly see democratic processes subjected to the same efficiency metrics that govern technological development, with calls to "streamline" decision-making, "automate" participation, and create "frictionless" civic engagement. The language itself reveals the problem. When we apply commercial optimization logic to democratic processes, we risk optimizing away the very characteristics that make democracy work.

Government is Not a Business; Democracy is not a Transaction

The mantra that "government should run like a business" has shaped public expectations and administrative practices for decades. This business-thinking substitutes concepts central to democratic governance. "Customers" replace citizens, "efficiency" trumps deliberation, and "streamlined service delivery" takes precedence over meaningful participation in collective decision-making.

While this thinking can improve certain institutional and administrative functions, it becomes problematic when applied to democratic exchange processes. Democracy is not a transaction; it is not a service delivery mechanism. It is a complex system for inquiry, imagination, contestation, and collective decision-making that necessarily involves disagreement, uncertainty, and the kind of messiness that business optimization seeks to eliminate.

AI accelerates this trend by making optimization appear not just preferable but inevitable. When technological solutions promise to make democratic processes faster and smoother, questioning these improvements can feel like standing in the way of progress. But this “progress” rhetoric obscures a crucial question: progress toward what? If the goal is more efficient democracy, AI may deliver. If the goal is more effective democracy – democracy that strengthens civic capacity, builds trust, enables meaningful participation, and enhances collective solidarity – the relationship becomes far more complex.