Q&A: California’s Top Politics Watchdog on AI and Influencers

The campaign business is at a crossroads.

Consultants are racing to adopt artificial intelligence tools into their tech stacks, candidates are increasingly courting online influencers and creators as campaign messengers and the media environment is more fragmented than ever. So far, federal regulators have been slow to address the shifting realities of campaigning in 2026. In their stead, states have moved to enact a patchwork of new laws and regulations in an effort to help shape the rules of this new era.

Campaigns & Elections spoke with California Fair Political Practices Commission Chair Adam Silver on Thursday about the new frontiers in campaign politics and how he’s thinking about the balance between free speech and regulation. 

The following transcript has been edited for length and clarity. 

C&E: The FPPC has been exploring AI tools of its own to flag irregularities in campaign reports. So I’m curious, where does the commission draw the line between AI-generated content that’s a normal part of modern campaigning and content that crosses into deceptive or actionable territory?

Adam Silver: That’s a really good question. I think, to a certain extent, since we’re a regulatory agency, our hands are tied. We have to follow what the intent of the legislature was in passing certain bills, including the AI disclaimer requirement. 

But they they gave us enough wiggle room that, based on the feedback that we received from all different types of stakeholders – whether that’s good government groups, campaign consultants, which we really relied on to get feedback on this, and of course the political attorneys who are helping those campaign consultants navigate all these requirements – we think we found a really good middle ground that is practically going to be easy to apply or straightforward to apply, but also is going to guarantee the public access to the level of disclosure they would expect. 

C&E: We’re seeing an explosion in these new technologies in campaigning. So how are you thinking about the balance between regulation and free speech here? Is this a new frontier, or is there anything that we can compare it to in the past? 

Silver: Yeah, it really complicates regulation in this area, but for good reason. We want to make sure that people can communicate and share their ideas and their concerns and their positions. 

I think, ultimately, in my personal opinion, there is nothing by its nature – by AI’s nature – that’s nefarious or bad or questionable, right? But I do think it’s important that where it is used to deceive the public – just like a regular Photoshop program would be used in the past – that the public receives the level of disclosure that they would expect; that it is disclosed, it’s identified as using AI, just because there is that that skepticism out there. 

I think one additional thing I’d flag there – and I actually went down this rabbit hole because of an article in your magazine – there’s this one campaign consultant in California, who, every time you all write an article on anything AI related, I get an email from him. So one of the emails he sent me related to research that the American Association of Political Consultants had done on the impact of AI disclaimers in campaign ads, and what they found – which I’m sure you already know – is that the inclusion of that disclaimer impacts how the recipient, the citizen, the voter interprets that message, or at least some of them. And that just further emphasizes how these sorts of rules that we’re interpreting and applying actually affect speech, because to a certain extent, that disclaimer is becoming an intermediary between the recipient and the speaker: the campaign. 

C&E: That’s a really interesting point to bring up. I saw the exact same study, and it found that, even if all the content is the same, if you put that disclaimer at the end saying the ad was generated by AI, it does reduce trust. I’ve talked to consultants who will say, “look, transparency is key here, and it is what it is.” Others will tell you that it’s costing credibility and preventing them from using this tool. What would you say to that?

Silver: I think in the campaign context, especially from my vantage point, transparency is the priority, right? But we also want to make sure that we’re considering the practical implications of rules that we’re adopting, promulgating, interpreting. Because ultimately, if we put rules out there that don’t really reflect what the desire of the public or the intent of the legislature is, you could end up undermining public confidence and disclosure. 

So one example would be: let’s say we ultimately decided to interpret an AI disclaimer extremely broadly. Rather than requiring campaigns and our enforcement team to make difficult decisions as to when this disclaimer should apply, let’s just apply it to every single advertisement that uses AI in any way. Then what you’re going to do is you’re just going to see this disclaimer on every single communication that goes out, and the public is potentially just going to disregard it. They’re just going to assume that AI is in control of the whole communications process for campaigns, and it’s going to have what I think would be a negative impact. 

Same sort of idea on the other side. If we interpret it particularly narrowly, then there’s a level of disclosure that I think the public’s not going to get. And if they see that the ethics transparency agency is furthering regulations that withhold some of that information or allow parties to withhold it, I think they’re going to have less trust in us, and certainly less trust in the parties. 

C&E: I want to turn to AB 1130 here. Obviously, it’s an issue for the legislature right now. But I’m curious, from your perspective, what’s driving the push for this bill, and how big a gap do you see between how much influencer-driven political content already exists and how much of it’s actually being disclosed?

Silver: That’s a good question. So for AB 1130, I do have an update for you. So you know we have appropriations committees in both houses that review all the legislation that’s going to have a fiscal impact. AB 1130 actually got through a pretty significant hurdle today, which was Senate Appropriations. Don’t quote me because I’m not our legislative director, but I suspect next it’ll be going to the floor. So we’re hoping that gets through relatively soon. 

Another thing to note there is we’re actually not the sponsor of that bill, though we strongly support it. The person who was behind getting that into a bill this session is the author, which is Assemblymember Marc Berman. And as I understand it, the reason for pursuing this so strongly that he would pursue it at this point in the legislative session is because he, just like the FPPC, identified an aspect of the law that undermined our ability to enforce it. 

So in current law, you have these disclaimer requirements on influencers that are paid to put out content supporting or opposing a measure or a specific candidate. However, the liability for violating that law is extremely limited, and we don’t have the ability to go after the influencers if they decide not to put the disclaimer on there, and we don’t have the ability to go after the committees if the disclaimer ultimately is not put on there. We only had jurisdiction over committees who fail to notify the influencers that they are required to put disclaimers on in certain circumstances. So that really took the teeth out of this really important requirement. 

If AB 1130 passes, then we’re going to have the ability to go ahead and prosecute folks who do not follow these rules, which, to be fair, is applicable in almost every other circumstance. If you’re paying a TV station, if you’re paying a mailhouse to put out mail, that disclaimer has to be on there. If it’s not, you can be fined for it. So I think that’s going to be the main reason for it. 

An aspect of it is also going to require more detailed reporting as to some of these payments because…some of these influencers have unique names that aren’t very familiar to the campaign community or the press. So I thought Mr. Berman’s office thought it’d be helpful to identify those sorts of payments to these third-party influencers as such, so that we can look to see whether disclaimers were provided and better enforce this law.  

C&E: Given how quickly campaign advertising is shifting towards digital and influencer channels, does the FPPC think California’s current disclosure framework can keep pace, or is more legislation like AB 1130 going to be a recurring necessity?

Silver: I think you’re always going to need to address advancements in the election campaign space through new legislation, just because strategies are evolving and they’re evolving quicker now than they ever have before. 

One thing that I’m personally concerned about is this concept called trend simulation. Are you familiar with trend simulation? Basically, how it works is people use the practice of clipping, where you pay a third party to put up a clip of, say, a longer speech or comment of some kind. I’m sure you see these on Twitter, X – whatever it’s called these days – all the time, right? So what folks are doing is they’re paying third-party agencies all at once to put out thousands of these clips on social media without any sort of indication that the clips are being paid for by a third party. And what it does is it creates the appearance of an organic trend. 

When TikTok or X or Instagram identifies that trend as organic, it’s going to start pushing that out to all of its users that it thinks it’s relevant to. So what you’re doing is you’re able to get specific content that you’re looking for your supporters, for voters to see, and it’s being sent out to the people that you would want to target for free, essentially, by using this practice. And just to be clear, there’s nothing wrong with clipping necessarily. But I think trend simulation is basically a deceptive practice in order to essentially trick people into thinking other people support you. 

So yeah, that’s a pretty scary thing that I think we’re going to need to learn more about, because it’s still kind of a relatively new practice, especially in the political space. I’m not aware of us coming across that sort of practice in the past, but it has become very common, almost ubiquitous, in the entertainment space.

C&E: For campaign professionals and consultants, what’s the one compliance blind spot around digital advertising or influencer partnerships that you would advise them to fix first before an enforcement matter. 

Silver: I would make sure that they’re educating their partners; they’re educating those influencers. If you’re going to go down that road where your campaign is going to pay third-party influencers to support your campaign or oppose your opponent, you need to make sure that third party understands what the rules are and understands that there are different rules in different jurisdictions. 

This is a very complicated world that a lot of unsophisticated parties are getting into without really thinking of the potential ramifications. So I think it’s important for campaign consultants to remind them so their clients don’t get hit with fines from their ethics agencies. 

C&E: Last question. States always have their own interests and concerns and their own dynamics, but at the same time, the Federal Election Commission has been without a quorum for over a year now. They’ve already punted on a rule-making process around AI. So, from where you’re sitting, is there more pressure on the states now to take a bigger role in overseeing federal campaigns?

Silver: That’s a good question. I mean, we don’t really have the ability to regulate federal campaigns, but I think what it really opens the door for us to do is really work with other jurisdictions who have adopted legislation and regulations in the context of AI disclaimers and in the context of third-party influencer disclaimers and come up with best practices and model language.

I’m sure campaign consultants…they’re complaining that there are so many different rules in different jurisdictions that it’s extremely difficult for their clients to use AI for clearly non-deceptive purposes, like creating stock photos and that sort of thing. So, what I think it does, I think it creates an opportunity for the jurisdictions who are doing this and doing it well to get together, identify what’s working, what’s not and create model language that other jurisdictions, when they’re pushed into this new area of the law, can come to us, or they could go to their stakeholders. And their stakeholders can say, “Hey, California does a great job of this. Michigan does a great job of this.” Because otherwise, I think a lot of times people are just looking to the feds for an answer here, and there really isn’t one.

C&E: Before we go, is there anything that you’d like to add?

Silver: I think one thing I’d want to share with campaign consultants is that ethics agencies aren’t as big boogeymen as they might think that they are, and deep down we really do want to collaborate with campaign consultants, because when we do, we can come up with better policy. 

There’s just certain things that a campaign consultant is going to know that we won’t, because we don’t have that experience, right? So I just think it’s critical that, if an agency reaches out to you for comment, even if you’re super busy, that you provide the relevant feedback. And if they’re not reaching out to you – I know everybody is extremely busy and there’s not even real off-election years anymore – you got to reach out to them. You got to make sure that they know you’re there, because then they’re more likely to collaborate with you. And I think that’s in everybody’s best interest: the public, the regulated community and the agencies. Hope that doesn’t come off preachy.