Americans Don't Trust What Government Does With Their Data — And They're Right to Ask

NewsOnScale Staff

October 28, 2026

There is a version of the civic technology story that gets told at conferences and in grant reports: government is modernizing, data is flowing, services are becoming more efficient, and democracy is being quietly upgraded in server rooms across the country. Then there is the version that shows up in public opinion research.

A new survey from the Center for Democracy and Technology finds that Americans are broadly worried about the personal information held by public agencies — and that this concern cuts across political affiliation, age, and geography. The anxiety isn't abstract. Respondents flagged specific concerns about how data is collected without clear consent, shared with third parties including law enforcement and private contractors, and stored without meaningful limits on retention or use.

These are not fringe concerns. They reflect a rational assessment of systems that have, in many documented cases, operated exactly the way people fear.

## The Gap Between Policy Language and Actual Practice

Government agencies have long described their data practices in terms of service delivery and public benefit. The forms you fill out to access benefits, the records created when you interact with a court or a school or a housing authority — these are framed as administrative necessities. What the CDT data surfaces is that the public increasingly understands there is a second life for that information, one that was never clearly disclosed and for which they never meaningfully consented.

That second life can include data sharing agreements with federal immigration enforcement, sale or licensing arrangements with data brokers, integration into algorithmic decision systems that affect credit, employment, and public benefits, and retention schedules that keep sensitive records active for decades.

The survey's finding that people want government accountability — not just privacy policies, but enforceable accountability — is significant. It suggests the public has moved past the point of wanting reassurance and is now asking for structural change.

## Why This Matters for the AI Agent Economy

This story is not only about legacy data practices. It sits directly at the intersection of the expanding AI agent economy and public infrastructure. Governments at every level are adopting AI-assisted tools for everything from benefits screening to predictive policing to child welfare assessments. Many of these tools are built by private vendors, trained on government-held data, and deployed under contracts that are partially or entirely shielded from public scrutiny.

When a resident interacts with a city's automated benefits portal, or when a social worker uses an algorithmic risk-scoring tool, the underlying data architecture is rarely visible to the person most affected by it. The CDT findings suggest the public senses this opacity even when they cannot name its technical components.

The accountability gap here is not hypothetical. Investigative reporting and litigation in multiple states have revealed cases where government AI tools produced biased or erroneous outcomes, where vendors retained data rights after contracts expired, and where audit mechanisms written into procurement agreements were never actually exercised.

## What Enforceable Accountability Would Actually Require

The survey's implicit demand — that accountability be real rather than rhetorical — points toward specific policy interventions that are technically feasible but politically difficult.

These include mandatory algorithmic impact assessments before deployment of AI tools in public-facing government functions, independent audit rights for civil society organizations, data minimization requirements that prohibit agencies from collecting information they cannot justify retaining, and private rights of action that allow individuals to challenge harmful data practices in court without depending solely on agency self-regulation.

None of these are novel ideas. Most have been proposed in various forms at the federal and state level. Few have passed in forms strong enough to matter.

## The Credibility Problem Is the Infrastructure Problem

What the CDT data ultimately describes is a credibility deficit. Public agencies are asking residents to trust systems those residents cannot see, governed by rules those residents cannot easily access, operated in partnerships with private entities those residents did not choose.

That deficit is not just a communications problem to be solved with better outreach. It reflects genuine structural features of how civic technology has been built and procured over the past two decades — optimized for administrative efficiency, with accountability treated as a compliance checkbox rather than a design requirement.

Rebuilding that trust will require more than transparency statements. It will require giving the public actual mechanisms to verify, challenge, and if necessary stop what is being done with their information.