Chapter 863 Says AI Is No “Person”
in Public-Participation Laws

Chapter 863 (SB 1159) was chaptered Sept. 30, 2026. The Legislative Counsel’s digest lists seven acts for which the bill specifies that AI systems are not a “person,” and the chaptered text says a person shall not knowingly use AI to falsely represent that a natural person engaged with an agency. Separately, the Bay Area Air District’s July 1, 2026 policy says emailed comments on regulations or plans will not be considered.

By William Douglas Baker

Founding Editor

Here, we read two California documents of 2026 as asking one question, whether a real person is behind a message to government: Chapter 863, chaptered Sept. 30, 2026, and the Bay Area Air District’s comment policy, dated July 1, 2026. The legislative status record shows that SB 1159 was chaptered by the Secretary of State as Chapter 863.

The July 1, 2026 staff memo lists, among a portal’s benefits, “greater confidence that comments are submitted by actual stakeholders rather than generated by an AI platform.” The policy says the Air District “will not consider or respond to electronic comments submitted by email or any other means outside of the public comment portal process.” According to the minutes of the Board’s July 1, 2026 meeting, as presented for approval on Sept. 2, 2026, a friendly amendment asked that the portal “not collect meta data or identifying information of users of the portal who opt not to share their information.” The portal’s page says “Your name and comments will appear on a public web page. Anonymous comments are also accepted.”

What Chapter 863 says

In Gov. Code section 7920.520(b)(1), Chapter 863 says that “person,” “interested person,” “member of the public” and any other similar terms for those who may engage with governmental agencies under the Public Records Act “do not include artificial intelligence systems, autonomous agents, robots, or other nonhuman entities, whether physical or digital.” The Legislative Counsel’s digest lists seven acts for which the bill specifies that “person” and similar terms do not include AI systems. Section 7920.520, as chaptered, says it “does not prohibit a natural person from using artificial intelligence, autonomous agents, or robots to facilitate the person’s own engagement with a governmental agency, including through the use of assistive technologies, provided the volume and frequency of the engagement are reasonably consistent with ordinary participation by a natural person.”

Gov. Code section 8319(a), as chaptered, says a person shall not knowingly use AI to represent falsely that a natural person engaged with an agency. As of October 22, 2026, the chaptered text names no Bay Area agency, vendor, council or newspaper.

The district’s policy

The policy says it governs electronic public comments on rulemaking and air quality plans. Among the concerns it lists is “the use of artificial intelligence to generate fraudulent comment emails that are not from real members of the public.” The second is the need to protect personally identifiable information in emailed comments, and the third is the potential for spam filters to keep emailed comments from reaching the intended staff recipient. It says its process will “provide greater confidence that comments received are from real members of the public.”

As of October 22, 2026, the district’s policy names no vendor, campaign, sender or case. The July 1, 2026 staff memo and presentation and the item-23 minutes name no Bay Area vendor, campaign or sender; the memo’s only named example of AI-generated comments is the South Coast district’s. None of the three says AI produced any Bay Area comment. The policy says the public may comment by non-electronic methods, such as a mailed or delivered letter during the specified comment period, or oral or written comments at a Board of Directors meeting.

Memo and vote

The memo says the increase in sophistication of artificial intelligence technologies “has enabled advocates to flood regulatory agencies with AI-generated comments.” The memo says the Los Angeles Times reported that the board of a different district, the South Coast Air Quality Management District, received over 20,000 emails. It was determined, the memo says, that many of the tens of thousands of email messages there were generated by an artificial intelligence platform used by an advocacy company. The memo says emailed public comments have on some occasions been filtered into spam folders and went unnoticed by staff until commenters followed up, and that email makes it “difficult to separate the content of the comments from Personally Identifiable Information (PII).” Staff proposes the portal, it says, “[i]n response to the potential for mass emails (including AI-generated ones) to mischaracterize stakeholders’ sentiments about Bay Area Air District regulatory and planning proposals and overwhelm staff capacity to properly process the comments.”

The memo says the portal would only allow the input of specific information from the commenter, including the commenters’ name, affiliation and contact information. According to the same minutes, the amended motion carried with 22 directors listed as voting yes, none no and two absent. Our comparison: the staff proposal had the portal include the commenters’ name; according to the minutes, the amendment asked that it not collect identifying information from users who opt not to share.

The Bay Area emails

The July 1, 2026 memo says “there is some indication that emailed comments recently received by Board Members on the Air District appliance rule amendment process were not sent by members of the public.” Its footnote cites a March 12, 2026 San Francisco Chronicle story. That sentence names no sender and does not mention AI. The Assembly committee’s analysis records the bill’s author saying that “the same playbook is being used to target the Bay Area air district, where a Chronicle investigation found that the residents who had comments submitted under their names never submitted anything.”

The Chronicle reported on March 12, 2026 that it reviewed 85 emails, which originated from domains ending in @Speak4.co, and that 10 people whose names were on the emails told it they did not send them or consent to Speak4 using their names. Rufus Jeffris, a spokesperson for the Bay Area Council, told the Chronicle: “We didn’t submit any letters on anyone’s behalf — this was all done by the users themselves.” The Chronicle wrote that “While Speak4 touts its ability to produce ‘AI-powered custom letters,’ Jeffris said that Speak4 did not use AI in its Bay Area campaign.” Speak4 did not respond to emails requesting comment, the Chronicle reported.

DeSmog reported on Oct. 5, 2026 that at least 85 of the letters sent to the air district “were sent by a software company that uses AI and was paid by an advocacy group called the Bay Area Council.” DeSmog also wrote: “The Council has denied using AI to send the forms.” No source this edition read shows, as of October 22, 2026, that artificial intelligence wrote any of the emails the Chronicle and DeSmog reported, or that an AI system sent them. The emails themselves, Speak4’s records and a transcript of the Assembly hearing of June 23, 2026 were not read. This edition does not say whether Chapter 863 reaches any campaign. The reasons it quotes for the district’s policy are the district’s own; it does not say whether any episode led to the policy.

Two tests

We read this as two different tests: Gov. Code 8319(a) turns on knowingly using AI to falsely represent that a natural person engaged with an agency, while the district’s policy turns on the channel a comment came through. Our comparison: the policy says emailed comments on regulations or plans subject to it will not be considered, and the portal’s page says anonymous comments are accepted.

Limits of this edition

This edition read the portal’s public page, not its submission form; it does not say what the form asks for, requires or checks, or whether the portal page’s wording follows from the Board’s amendment. Draft minutes of the Board’s Sept. 2, 2026 meeting, in its Oct. 7, 2026 packet, record the Board approving the July 1 minutes on its consent calendar; this edition read the July 1 minutes as presented for approval.

The Bay Area emails enter this edition through two official records, the Assembly committee’s analysis and the district’s July 1 staff memo, and through the news reports quoted above. Ours: we tag this record, not any person or company, to Trust Transfer Failure, here in reverse and loosely, within Digital Derangement Syndrome™ (DDS™), the house name for the condition in which AI systems fail to recognize real expertise; Chapter 863 and the district’s policy both ask whether a real person is behind a message to government.

Before an agent speaks under your name, ask each institution which question it asks: what made the message, or which door it came through.