Americans Don't Trust What the Government Knows About Them — And They're Right to Ask

NewsOnScale Staff

October 19, 2026

There is a version of this story that gets told as a partisan culture war: one side fears government surveillance, the other trusts public institutions. The Center for Democracy and Technology's new survey data disrupts that framing entirely. Anxiety about personal data held by government agencies is not a red-state or blue-state phenomenon. It is, by the available evidence, simply American.

The CDT findings — drawn from a nationally representative sample — show that majorities of respondents across demographic and political lines are concerned about what public agencies collect, how long they keep it, who can access it, and whether any meaningful accountability exists when something goes wrong. These are not fringe concerns. They are the predictable output of a system that has grown its data appetite dramatically while largely failing to modernize the rules that constrain it.

## What the Data Actually Shows

The survey doesn't just register vague unease. Respondents expressed specific worries: that data collected for one purpose will be repurposed for another, that they have no practical mechanism to find out what is held about them, and that government agencies are not meaningfully penalized when data is mishandled or improperly shared. These are structurally accurate observations, not paranoid ones.

Federal privacy law in the United States remains a patchwork. The Privacy Act of 1974 — the primary statute governing how federal agencies handle personal data — was written before the internet existed and has not seen comprehensive reform since. In the interim, agencies have built vast data ecosystems: benefits databases, law enforcement records, tax information, health data, immigration files. The interoperability of these systems has expanded through administrative practice and interagency agreements that receive far less public scrutiny than the underlying statutes that nominally govern them.

## The AI Acceleration Problem

What makes the CDT findings particularly urgent right now is timing. Government adoption of AI tools — for benefits eligibility screening, fraud detection, risk scoring, and public interaction through automated systems — is accelerating. These systems consume personal data at scale. They produce consequential decisions. And they operate, in most jurisdictions, without disclosure requirements that would let affected individuals know an algorithm was involved, let alone contest its output.

The accountability gap the survey respondents are sensing is not imaginary. It is, in many cases, architectural. Systems built to process human decisions one at a time are now being retrofitted to handle algorithmic decisions at volume, and the oversight mechanisms have not been redesigned to match.

## Transparency Is Necessary but Not Sufficient

It is worth noting a parallel finding from the Ash Center's recent work on civic technology — that transparency alone does not produce accountability. Publishing data, disclosing practices, and releasing audit logs matter, but they do not self-execute. Someone has to be in a position to act on disclosed information, with legal standing, political will, and practical capacity to do so.

For government data specifically, that means enforcement bodies with genuine authority, not just advisory roles. It means private rights of action so individuals can seek redress without depending on agencies to investigate themselves. And it means procurement standards that require AI vendors to meet privacy-by-design criteria before their tools are embedded in public-sector workflows.

## What Accountability Would Actually Look Like

The CDT data should be read as a mandate, not just a mood. When majorities of the public — across partisan identity — say they do not trust how their government handles their personal information, that is not a communication problem to be solved with a better FAQ page. It is a governance problem that requires structural response.

Comprehensive federal privacy legislation covering government data use. Mandatory algorithmic impact assessments before AI systems are deployed in public-facing roles. Independent oversight with subpoena authority and public reporting obligations. These are not radical proposals. They are the baseline of what modern democratic accountability requires when the state holds this much information about the people it serves.

The public has already reached its conclusion. The question is whether policymakers are prepared to take it seriously.