A Southern California air quality agency faced a potential case of automated comment fraud when it received 20,000 opposition comments to a proposed heat pump rule—far more than typical public responses. The South Coast Air Quality Management District became suspicious when the volume and nature of submissions seemed inauthentic, including an email thanking the executive director for opposing a rule his own team had drafted.
When the agency contacted a sample of 172 commenters to verify authenticity, almost no one responded. Of the five people who did reply, three said they knew nothing about comments submitted under their names. A separate Sierra Club investigation found similar results, with four contacted individuals also denying they had sent emails.
According to the L.A. Times, CiviClick, a company offering "AI-powered advocacy tools," led the opposition campaign on behalf of a public affairs consultant with gas industry ties. CiviClick denies sending emails without consent or using AI to fabricate automated messages, while the air quality district continues investigating and exploring more aggressive verification methods.
The incident highlights growing concerns about AI's potential to manipulate democratic processes through fake public comments. Government agencies now face the challenge of distinguishing genuine citizen input from automated submissions, as AI technology makes it increasingly easy to impersonate humans and submit comments without individuals' knowledge.