|
AI Agents Submitted 20K Public Comments in LA: Where is Impactful Engagement? I've been talking about this for years. Public engagement processes don't always reflect the public. The recent outrageous Los Angeles incident should change that. Over 20,000 comments flooded the South Coast Air Quality Management District opposing proposed rules to phase out gas appliances. Many were reportedly generated through an AI platform called CiviClick. Some people listed as commenters say they never submitted anything (Los Angeles Times, 2026). The board voted against the rules 7-5. Everyone is talking about the AI. But we should pause on something more uncomfortable: we already knew this was happening. Organized, well-funded interests have always found ways to dominate public comment processes through coordinated letter campaigns, paid consultants, and manufactured grassroots opposition. And,. we still called it community engagement and moved on. What's different now is that AI isn't just helping humans send more messages. AI is acting as an agent (e.g. by identifying audiences, crafting personalized letters, deploying them at scale across multiple channels, and doing it faster and cheaper than any human campaign ever could). This is not a new issue. AI just magnified it. And, I hope it forces all of us to move away from check-boxing engagement approaches. I hope educators consider reality of implementing impactful public engagement processes when teaching students and the future workforce about public engagement.
This incident didn't introduce a new flaw into public engagement. It held a mirror up to one we've been living with for a long time. The scale (Gladly) made it impossible to dismiss. Some of the named commenters didn't even know their voices had been used. So let's use this moment to reflect. Not just on AI, but on what we actually want public engagement to do. What’s its real intended impact? Clarity: Do people genuinely understand what's being decided? The AQMD rules had been nearly two years in the making, projected to prevent an estimated 2,500 premature deaths and over 10,000 new asthma cases. Yet many who opposed them described the proposal as a mandate (which by the way officials said wasn't accurate). When people lack clear information, they become easy to mobilize through fear. Clarity is a defense against manipulation (by any tool, human or AI agents). Trust and power: When a well-resourced campaign can flood an agency with thousands of comments , it doesn't just skew an outcome. It corrodes trust in the entire process. And we all know that trust, once lost, is very hard to rebuild. Trade-offs: Sometimes the evidence already tells us what will help a community. The question we rarely ask out loud is this: if we know engagement will be dominated by organized interests, what obligation do we have to the people who couldn't show up? AI acting as an agent in civic life is new. But the problem it's amplifying is not. This is a good moment to stop treating public engagement as a box to check and start asking whether it's actually doing what we always claimed it was for.
|
