Most Americans Don't Trust Government With Their Data — And They Have Good Reason Not To

NewsOnScale Staff

September 11, 2026

There is a quiet crisis of legitimacy unfolding inside the bureaucratic machinery of American government — not in the halls of Congress or on cable news, but in the servers, databases, and algorithmic systems that process the most intimate details of millions of people's lives.

A new survey from the Center for Democracy and Technology (CDT) makes the scope of public concern concrete: Americans, across demographic and political lines, are worried about the personal data held by public agencies, and they want something done about it. The findings are notable not because anxiety about government surveillance is new, but because the survey captures how that anxiety has sharpened and broadened as digital government services have expanded — and as AI-powered decision systems have begun influencing everything from benefit eligibility to policing.

## What the Public Actually Thinks

The CDT data cuts against a convenient assumption often made in govtech circles: that ordinary people are either indifferent to data practices or willing to trade privacy for convenience. That turns out not to be true. Majorities of respondents expressed discomfort with government agencies sharing personal data across departments without explicit consent, with the use of that data in automated decision-making, and with the absence of clear mechanisms to challenge or correct government records about them.

Perhaps most striking is the demand for accountability infrastructure — not just vague calls for "more regulation," but specific appetite for audit rights, transparency reports, and independent oversight of how agencies deploy data systems. This is a public that has, slowly and often through bad experiences, developed a more sophisticated understanding of what data collection actually means in practice.

## The Accountability Gap Is Real

The concern is well-founded. Across federal agencies and state governments, the adoption of data-intensive and AI-assisted systems has outpaced the development of any coherent accountability framework. Agencies use third-party data brokers to supplement what they collect directly. Predictive tools inform child welfare investigations, parole decisions, and fraud detection — often with minimal transparency about how those tools work or how to contest their outputs. Records are retained for years or decades under rules designed for paper files, not persistent digital profiles.

The legal scaffolding governing all of this — the Privacy Act of 1974 being the primary federal instrument — was built for an entirely different technological era. Reforms have been incremental and uneven. A handful of states have passed their own data protection laws, but coverage is patchy and enforcement is frequently underfunded.

Meanwhile, the govtech industry continues to market AI-powered solutions to agencies hungry for efficiency gains, often with procurement processes that lack the technical expertise to evaluate what those systems actually do or what risks they introduce.

## Why This Matters for the AI Agent Economy

For those watching the broader AI agent economy, the CDT findings point to a structural tension that will only intensify. As government services become more automated — with AI systems making or heavily influencing decisions about housing assistance, tax compliance, immigration status, and healthcare access — the question of who is accountable when those systems fail or discriminate becomes urgent and largely unanswered.

The same dynamics that make algorithmic systems appealing to administrators (scale, speed, apparent objectivity) are precisely what makes them dangerous without robust oversight. An AI that wrongly flags a benefits recipient for fraud can cause immediate, concrete harm. If there is no clear appeal process, no explainability requirement, and no public audit trail, the affected person has almost no recourse.

## What Accountability Would Actually Look Like

The CDT survey implicitly sketches an accountability agenda: mandatory transparency reporting from agencies deploying automated decision tools, individual rights to access and correct government-held data, independent technical audits of high-stakes algorithmic systems, and meaningful civil penalties for agencies that misuse or inadequately protect personal information.

None of this is technically complicated. It is politically complicated — which is a different problem, and one that public pressure can, over time, move.

The fact that Americans are now articulating these demands with some precision is itself significant. Distrust without direction produces cynicism. Distrust with a concrete accountability agenda produces policy pressure. The CDT data suggests the public may be closer to the latter than Washington has recognized.