campaigns

Contributor: Investigate the AI campaigns flooding public agencies with fake comments

California built its tradition of open government — including for citizen boards that set the rules for such functions as automotive repair and security guard licensing — precisely to keep well-funded corporate interests in check. Lobbyists and special interests are constantly scheming to defeat the will of the majority. Now they are able to do more damage using artificial intelligence to simulate fake grassroots opposition to clean air measures, and they are surreptitiously using the identities of real people to deceive regulators.

Last June, the South Coast Air Quality Management District received more than 20,000 comments opposing a pair of clean air rules that would have prevented 2,500 premature deaths and 10,000 new cases of asthma. A February investigation by the Los Angeles Times revealed that those comments were submitted through CiviClick, a Washington-based AI-powered comment generation platform, orchestrated by a local political consultant with ties to the natural gas industry. When the district’s cybersecurity team reached out to a small sample of commenters to verify their identities, a majority of respondents said that they had not submitted the comments in their names.

Even so, the flood of fake comments seemingly worked. These rules, vehemently opposed by the natural gas industry, already watered down by the district to near-toothlessness, were ultimately rejected by the board — apparently overwhelmed by the flood of fake opposition to even the mildest effort to limit pollution from gas-burning appliances.

This Southern California campaign was not an isolated incident. A recent investigation by the San Francisco Chronicle also revealed that an industry front group used Speak4, a platform that advertises its use of AI, to submit dozens of comments regurgitating talking points from the fossil fuel industry in an attempt to weaken and delay clean air rules in the Bay Area. The scheme was exposed when 10 residents whose identities were used on these emails said they absolutely did not send them, calling the messages “forged.”

In both cases, organizations submitted emails and comments to regulators using real people’s identities without their knowledge or consent. This playbook has been employed in other states: CiviClick was used by fossil fuel companies to support a gas-pipeline-expansion project in North Carolina last year. When elected officials reached out to a few respondents to verify the messages, some constituents stated they had no knowledge of the emails sent under their names.

The opposition campaign to South Coast’s clean air rules was run by one of the state’s most powerful lobbying firms. Its client list includes Sempra, the parent company of SoCalGas, which opposed the clean air standards, which would have encouraged the sale of pollution-free heat pumps and threatened the utility’s business.

The industry front group using AI to undermine clean air rules in the Bay Area, Common Sense Coalition, also has ties to fossil fuel companies. Common Sense Coalition is a project of the Bay Area Council, a local business group that features members such as the Western States Petroleum Assn., Chevron, Martinez Refining Co. and Phillips 66.

The question of whether fossil fuel interests financed astroturf AI campaigns to defeat clean air rules should be answered through full investigations, which also ought to address whether the campaigns committed fraud and identity theft.

Californians deserve to know what is going on — how AI was used, where the lobbyists got the names and addresses they attached to the robo-messages and who paid for the deceptive campaigns. What’s most concerning is the use of actual residents’ identities — without their knowledge or consent — to oppose life-saving clean air standards.

Top law enforcement officials should be investigating — including Atty. Gen. Rob Bonta, Los Angeles Dist. Atty. Nathan Hochman and San Francisco Dist. Atty. Brooke Jenkins.
If the law on using a person’s name in a scheme to thwart action by a public agency is not clear enough to support prosecutions, then the law needs to be tightened up — and there is legislation, Senate Bill 1159, aiming to do that.

If this seems like a niche issue, I can assure you it is not. I spent 17 years at the helm of the California Air Resources Board, and I am deeply disturbed by the potential co-opting of public input processes using forgery through automated tools. Gathering public input is fundamental to the legitimacy of regulatory agencies.

We frequently heard from individuals or business associations concerned about the cost or burden of proposed regulation, and we worked hard to understand and tailor our rules to make them as streamlined and cost-effective as we could, while still making progress toward reducing the air and climate harms of a wide array of equipment and activities.

The destruction of meaningful public input through deceit isn’t just an environmental issue; it’s a democracy issue — and it demands urgent attention and accountability. California should draw the line to protect our democratic institutions.

Mary Nichols was chair of the California Air Resources Board, where she occupied the attorney seat. She is distinguished counsel to the Emmett Institute on Climate and Sustainability at UCLA Law School.

Source link