The Counterfeit Citizen

The Automated Public Square

Modern regulatory governance relies heavily on notice-and-comment procedures. When executive agencies propose rules governing environmental standards, labor protections, or corporate mergers, federal and state laws require them to open public dockets. This mechanism was designed to give ordinary citizens a direct voice in executive decision-making.

That channel is now being systematically degraded. Advanced generative systems have transformed public comment processes into targets for automated astroturfing. Where political campaigns once relied on form letters or manual outreach to demonstrate grassroots interest, interest groups can now manufacture thousands of unique, persuasive constituent letters in minutes.

The threat is not subtle propaganda; it is synthetic constituency. By flooding regulatory dockets with artificially generated arguments, well-resourced advocacy groups and corporate interests can fabricate the appearance of widespread public sentiment. The foundation of participatory rulemaking is dissolving under an avalanche of machine-written civic input.

Agentic Form-Filling and Synthetic Constituency

Public concern over artificial intelligence in politics has focused primarily on election deepfakes and social media disinformation. However, the operational corruption of government rulemaking happens through a distinct mechanism: agentic form-filling at scale.

Autonomous software agents are configured to navigate government portals, complete mandatory registration fields, and submit tailored comments across multiple administrative dockets. Rather than submitting identical text strings that simple spam filters could easily flag, generative language models alter sentence structure, vocabulary, tone, and personal narrative across thousands of submissions.

A single operator can instruct an agentic workflow to generate ten thousand distinct comments supporting or opposing a proposed rule. One submission presents the perspective of a concerned small-business owner; another adopts the tone of a retired civil servant; a third emphasizes economic anxiety. Each argument sounds authentic, yet none corresponds to a human constituent.

The economics of political influence have changed permanently. Industrial-scale public lobbying no longer requires mobilising real voters or hiring extensive campaign staff. Synthetic advocacy lowers the marginal cost of manufacturing public consensus to pennies per docket entry.

The Failure of Volume as a Proxy for Democracy

Regulatory agencies face an operational crisis. Under administrative law, agencies are not legally required to count public comments as votes, but volume has long served as an informal proxy for public intensity and interest. When a proposed rule receives hundreds of thousands of comments, officials take notice.

Synthetic comment flooding renders volume meaningless. Agency analysts responsible for reviewing dockets cannot distinguish between organic community outrage and targeted algorithmic execution. When fifty thousand comments arrive on a technical rule, staff members must spend months cataloging submissions that may contain zero human input.

This dynamic creates an artificial consensus illusion. Powerful interest groups can artificially inflate public opposition to environmental regulations or consumer protections, making balanced rules appear politically contentious. Regulators, fearful of litigation or political backlash, hesitate to act on clear statutory mandates.

The distortion extends beyond federal rulemaking. Local city councils, public utility commissions, and regional planning boards face identical automated campaigns. As local government portals adopt digital intake forms, synthetic constituency spreads into municipal governance.

The Asymmetric Detection Race

Faced with automated comment campaigns, government agencies have sought technical remedies. Officials have turned to automated text classifiers and machine-learning detection algorithms to identify AI-generated submissions. These efforts face severe technical constraints.

Text detection algorithms suffer from high false-positive rates, frequently misidentifying genuine human writing as machine-generated text. Non-native speakers, citizens using standardized templates, and individuals relying on basic writing assistance are disproportionately misclassified. Relying on automated detection risks disenfranchising authentic human participants.

Conversely, state-of-the-art language models easily evade detection metrics. By introducing stylistic variations, subtle grammatical imperfections, or localized jargon, generative agents produce text that bypasses standard statistical checks. Generation capabilities advance exponentially, while detection accuracy remains brittle and reactive.

Implementing heavier friction, such as mandatory identity verification or digital credentials, creates new barriers to democratic participation. Requiring government identification or complex authentication deters marginalized communities while offering minimal deterrence to sophisticated political actors capable of harvesting public registration data.

The Retreat to Technocracy

The ultimate damage of synthetic comment flooding is institutional disillusionment. When administrative dockets become unusable as measures of public sentiment, government officials inevitably retreat from public consultation altogether.

Faced with compromised public dockets, regulatory bodies increasingly rely on closed expert panels, private industry consultations, and informal stakeholder meetings. Decision-making moves out of transparent public dockets and back into closed rooms inaccessible to ordinary citizens. The public consultation process becomes a formal exercise rather than a meaningful dialogue.

This shift alienates the citizenry from administrative governance. Real human beings who take time to draft thoughtful comments find their voices buried beneath thousands of synthetic submissions. When citizens realize their genuine feedback is indistinguishable from machine output, civic engagement declines.

The collapse of trust flows in both directions. Officials view public dockets with cynicism, while citizens view administrative processes as rigged. Synthetic intelligence does not need to convince anyone of a lie to undermine democracy; it only needs to render civic channels useless.

Verification or Decoration

Democratic legitimacy rests on the principle that public governance responds to human needs, values, and experiences. If administrative consultation cannot distinguish between authentic human participation and automated prompt engineering, public comment ceases to be democratic oversight.

Preserving participatory rulemaking requires establishing clear standards for verified human engagement. Regulatory systems must develop privacy-preserving methods to confirm that public submissions originate from real individuals, without creating exclusionary barriers for disadvantaged communities.

Without robust standards for human authenticity, public consultation degenerates into political decoration. Agencies will continue hosting comment portals to fulfill statutory mandates, while ignoring the content submitted to them. A democracy that cannot verify the humanity of its participants will soon cease to care what they have to say.

Source Note

This analysis draws on administrative law research regarding notice-and-comment rulemaking, studies on digital astroturfing techniques, policy reports on agency docket management, and evaluations of text-detection reliability. Claims regarding regulatory vulnerabilities reflect structural systemic dynamics rather than specific active dockets.