Media Advisory - FSF Files Comments on FCC's Propose Rules for AI Generated Content in Political Ads

Media Advisory

September 19, 2024

Contact: info@freestatefoundation.org

Free State Foundation President Randolph May and Seth Cooper, Director of Policy Studies and Senior Fellow, submitted comments today in the Federal Communications Commission’s proceeding proposing to require radio and TV broadcasters as well as cable and direct broadcast satellite (DBS) operators to include a disclaimer on all political ads that contain content generated by artificial intelligence (AI). These comments demonstrate that the Commission lacks statutory authority to adopt its proposed regulation of the content of political ads using AI and that, in any event, it would constitute unsound policy to do so.

The complete set of the Free State Foundation comments, with footnotes, is here.

 

Immediately below are the "Introduction and Summary" to the comments, without the footnotes.

Introduction and Summary

These comments are submitted in response to the Commission’s Notice proposing to require radio and TV broadcasters as well as cable and direct broadcast satellite (DBS) operators to include a disclaimer on all political ads that contain content generated by artificial intelligence (AI). They also would be required to include a notice in their online political files disclosing the ad’s use of AI. The Commission’s rush to adopt a novel AI political ad regulation is a misguided power grab – a combination of bad law and bad policy. The Commission should not adopt the proposed rule.

 

The agency lacks statutory authority for its proposed regulation of the content of political ads using AI. The Notice of Proposed Rulemaking cites Section 303(r) and other provisions of Title III of the Communications Act regarding the agency’s power to make rules and regulations necessary to carry out the Act’s provisions in the “public interest.” But the Commission has no traditional regulatory authority over the content of political ads on broadcast radio or TV, and none of those provisions cited in the Notice contain language that reasonably may be interpreted to authorize disclaimer and disclosure mandates for political ads featuring AI-generated content.

Moreover, the FCC’s proposal is likely to run afoul of the Major Questions Doctrine (MQD) as articulated in West Virginia v. EPA (2022) because it involves a question of “vast economic and political significance.” Proposing for the first time to regulate the use of AI in connection with political advertisements appears to be a paradigmatic case meeting the MQD criteria. As such, and because Congress has not clearly granted the FCC authority to adopt the rule it proposes, it’s very unlikely to survive judicial review.

 

By contrast, the Federal Elections Commission (FEC) is given much more explicit statutory authority to regulate significant aspects of political campaign ads under the Federal Election Campaign Act. This includes the FEC’s “exclusive jurisdiction with respect to the civil enforcement” of the Act. To date, however, the FEC has never determined it has jurisdiction to regulate political ads with AI-generated content under its “materially deceptive” statute – and the FEC may lack such authority. If the FEC lacks authority to regulate political ads with AI-generated content, then a fortiori the FCC certainly lacks similar authority under Communications Act provisions regarding broadcast, cable, and satellite services. 

Even if the FCC had the requisite legal authority, the proposal constitutes bad policy because it would apply to ads with AI-generated content that are not materially deceptive, likely causing many viewers to distrust the ads solely or primarily because of the boilerplate disclaimer or simply to “tune out” the disclaimers. Also, it would apply only to ads that are broadcast or transmitted by FCC-regulated services – and not by Internet outlets that garner an increasing share of political ads. Requiring disclaimers on ads shown by broadcast, cable, and satellite services when those same ads may be posted online to wider audiences without disclaimers will add to the confusion, especially since materially deceptive ads are more likely to appear online. Moreover, broadcasters (and cable and DBS operators) do not have inside knowledge about how given political ads were created; yet under the proposed regulation, apparently they would shoulder the burden of having to discern when generative AI was used. By focusing on broadcasters of political ads rather than the creators, the proposed regulation deviates from a more reasonable focus on ad creators that is taken in many nascent state laws regulating the use of AI in elections.

 

Additionally, the proposal would put the Commission in the untenable position of making judgments about “credible third parties” who raise complaints about ads, a matter in which the agency has no expertise. Government should not assume any role in designating third parties as “credible” or not credible for purposes of deciding whether political ads should be disclaimed, disclosed, or taken down. If it were to do so, it would inevitably, and justifiably, invite suspicion that its decisions are politically motivated. The proposed overly broad definition of “AI-generated content” likely would result in broadcast, cable, and satellite services requiring disclaimers for all or nearly all political ads as a regulatory risk aversion measure, rendering such disclaimers unhelpful, if not meaningless.

A PDF of the complete set of Free State Foundation comments, with footnotes, is here.

Thinking Things Through XI: On Weed Whackers, Partisanship, and Regulatory Philosophy

I expect the Federal Communications Commission under new Chairman Ajit Pai to do well by getting the agency to do less. But before the agency can do less, in one sense it has to do more. That is, more in the way of curtailing existing regulations that either shouldn’t have been adopted in the first place or that are no longer necessary or cost-justified.

At his address at the Free State Foundation’s Tenth Annual Gala Conference on December 7, 2016, Chairman Pai made his deregulatory intent clear, stating that, “[i][n the months to come, we also need to remove outdated and unnecessary regulations.” Perhaps with only a bit of hyperbole, he declared: “We need to fire up the weed whacker and remove those rules that are holding back investment, innovation, and job creation.” I’m pleased that he added immediately thereafter: “Free State and others have already identified many that should go.”

Chairman Pai is off to a good start. He’s pulled from current consideration two significant draft orders prepared under former Chairman Tom Wheeler’s direction. One would have imposed ill-conceived, costly regulatory mandates on TV set-top boxes at a time when the video marketplace is undergoing rapid change in the direction of increased consumer choice and competition. And, in the process, the proposed regulation would have undermined the ability of content creators to protect their copyrights and their contractual interests in video programs.

The other draft order would have re-regulated the rates – in effect, forced them lower – for business broadband services offered by the former “telephone” companies. This would have come at a time when new entrants like former “cable” companies increasingly are gaining a competitive foothold. In other words, the Wheeler proposal would have made it more difficult for new facilities-based (emphasis on facilities-based!) entrants like cable operators to compete, hindering the development of a more competitive market.

Chairman Pai also withdrew a draft “Section 706” report prepared under Mr. Wheeler’s direction. Under Chairman Wheeler, these broadband deployment status reports regularly had been used as a means to the end of justifying more broadband regulation. No matter how speedy the pace of deployment, or the increase in available bandwidth, the Wheeler Commission refused to determine that deployment is occurring, in the words of the statute, “in a reasonable and timely fashion.” The goalposts kept moving – so that the Commission somehow could avoid finding broadband deployment reasonable. I suspect that under Chairman Pai the Section 706 reports will be grounded in marketplace realities rather than in preconceived notions.

*     *     *

Now, all that said, I want to suggest this important point regarding regulation and regulatory agendas at the FCC. It would be a mistake, and naïve, to assume, as some are suggesting, that there won’t any longer be partisan splits on votes on major items under a Pai Commission. I expect there will be – and there is nothing inherently wrong with this to the extent that commissioners’ votes are grounded in differing philosophical principles or perspectives. After all, there are real differences in philosophical perspective between the parties with which the commissioners identify regarding the proper role of government regulation.

It is blinking reality to suggest communications policy is somehow immune, or should be, from application of different regulatory perspectives that, at least at times, are rooted in different partisan affiliations. Especially with regard to formulating communications policies, consistent with congressional delegations of authority, that are dependent on criteria such as the extent of existing and potential competition, the rate of technological change, the forecast evolution of consumer demands, and so forth, there are no ironclad Newtonian laws that govern, no E=MC². Many times the “right” answer in evaluating competition or competitive impact, for example, won’t be found in some mathematical equation, not even a Herfindahl-Hirschman Index.

The FCC is established as a bipartisan agency, not a nonpartisan one. Unlike the Federal Election Commission, for example, which by law must be composed of an equal number of Democrats and Republicans, the five-member FCC may have no more than three members from the same party. And the President designates the FCC chairman, so, as former President Obama famously reminded us early in his first term, elections have consequences. There is no reason not to acknowledge forthrightly that Ajit Pai and Michael O’Rielly adhere to a more deregulatory perspective than the more pro-regulatory one of, say, Tom Wheeler, Mignon Clyburn, and Jessica Rosenworcel.     

Now, I do not mean to suggest for a moment that efforts should not be made at the FCC to minimize partisan differences, especially on contentious items. I expect that Chairman Pai will make such efforts to seek compromises where possible. Just engaging in such efforts often can lead to sounder decisions. After all, a multimember commission like the FCC is grounded, at least in part, in the idea that good-faith collaborative discussions among commissioners with a genuine receptivity to considering opposing viewpoints will produce better decisions, even if not unanimity.

*     *     *

Finally, in acknowledging the legitimate role that differing philosophical perspectives necessarily play in the realm of formulating communications law and policy, I don’t mean to suggest – and don’t want to be misunderstood as suggesting – that in reaching particular decisions, whether in the context of adjudications or rulemakings, careful consideration of actual record evidence is not of utmost importance. Of course, it is. And any FCC commissioner who fails to accord a proceeding’s evidence proper consideration is derelict.

But it serves no useful purpose to deny that when decision-time comes, the conclusions to be drawn from the evidentiary record may differ. Again, particularly when it comes to matters based on judgments regarding the extent of existing market competition and future prospects, or predictions concerning the rapidity of technological advances, or of prospective consumer benefits or harms, there are no mathematical formulas to apply.

It is in this context, and especially, say, when the evidence is roughly in equipoise, or seems to be – or as the economists say, “well, on the one hand this, but on the other that” – that it is perfectly proper for a commissioner with a deregulatory orientation, as a matter of philosophical principle, to resort to a default presumption in favor of not imposing a regulatory mandate. In other words, as I have recently proposed, in the absence of clear and convincing evidence to the contrary, and within the confines of the Communications Act, in today’s Digital Age marketplace environment in which competition and consumer choice have become increasingly ubiquitous, employing a deregulatory presumption is entirely proper.

In sum, Chairman Pai’s regulatory “weed whacker” surely has its uses. Otherwise, as he put it in his Free State Foundation address, unnecessary regulations will hold back “investment, innovation, and job creation.”

FCC GON Wild

Last week, briefs were filed in the court of appeals by parties challenging the FCC’s preemption of a Tennessee law imposing restrictions on a local government’s ownership and operation of a broadband network. In the same order, the Commission also preempted a similar North Carolina law. In FCC acronym-land, a government-owned network is commonly referred to as a GON.

As in … FCC GON wild.

I have never suggested that, as a matter of policy, GONS should be prohibited in all circumstances at all times. If there are particular areas that private sector broadband providers simply are not serving, and do not intend to serve, then a GONS, or some form of public-private partnership, may be appropriate. But these cases, by far, should be the exception, not the rule.

As my colleague Seth Copper and I explained in a May 2015 article, FCC Preemption of State Restrictions on Government-owned Networks: An Affront to Federalism, in the Federalist Society publication Engage:

A threshold issue is the problematic nature of government assuming the dual role of both enforcer of public law and competitor to private sector providers. This duality poses inherent conflicts-of-interest. For example, local governments may excuse their own networks from running the bureaucratic permitting and licensing gauntlet through which private providers must pass. Fear of disfavored treatment deters private market investment in broadband infrastructure. In addition, questions concerning the institutional incentives and competency of local governments operating capital-intensive advanced communications networks in rapidly innovating markets heighten the concerns of local taxpayers. And speech restrictions that are common in the terms of services of government-owned networks raise significant First Amendment issues.

So GONS are problematic from a policy perspective. But I want to use the occasion of the filing of the initial appellate briefs to emphasize the highly questionable nature of the FCC’s preemptive action as a matter of law. The FCC’s action raises rule of law concerns that are at the heart of our federalist constitutional system.

Our Federalist Society article focused primarily on theses serious legal issues. To my mind, the FCC’s preemption order is sufficiently beyond the authority delegated to the agency by Congress, and beyond the bounds of the Constitution’s federalist structure, that I suggest this is a case of the FCC GON wild. Below are a few excerpts from the article that highlight why the Commission’s action is likely to be overturned in court.

*   *   *

“The most obvious difficulty with basing preemptive authority on Section 706 is that the statute’s language nowhere authorizes it. Section 706(a) provides:

The Commission and each State commission with regulatory jurisdiction over telecommunications services shall encourage the deployment on a reasonable and timely basis of advanced telecommunications capability to all Americans...by utilizing, in a manner consistent with the public interest, convenience and necessity, price cap regulation, regulatory forbearance, measures that promote competition in the local telecommunications market, or other regulating methods that remove barriers to infrastructure investment.

Preemption is not one of the enumerated measures or methods. Inferring preemption from Section 706(a) is also difficult because of its poor fit with the statutory structure. Section 706(a) recognizes a role both for ‘[t]he Commission and each State commission with regulatory jurisdiction over telecommunications services.’ Federal preemption of state laws imposing geographic or other forms of restrictions or safeguards on government ownership of broadband networks disregards the role of state officials that the statute explicitly acknowledges.”

*   *   *

“In a 1997 order, the FCC rejected a petition requesting it to preempt state law restrictions on municipal telecommunications networks based on Section 253(a) of the Communications Act. … As the FCC’s 1997 order declared: ‘[S]tates maintain authority to determine, as an initial matter, whether or to what extent their political subdivisions may engage in proprietary activities.’ It also observed that preemption ‘effectively would prevent states from prohibiting their political subdivisions from providing telecommunications services, despite the fact that states could limit the authority of their political subdivisions in all other respects.’

This agency precedent cannot be avoided simply because Section 706 is now invoked as opposed to Section 253. The states’ authority to decide ‘whether or to what extent their political subdivisions may engage in proprietary activities’ is not altered just because a particular FCC majority wants local governments to offer broadband services. Federalism principles previously recognized by the FCC, grounded in the Constitution, do not lend themselves to dismissals based on ‘reasonable explanations’ about current Commission policy objectives. For that matter, the 1997 Order recommended states consider restrictions on government-owned networks rather than totals bans. The FCC’s present about-face regarding such restrictions hardly seems reasonable. Indeed, it seems arbitrary and capricious.”

*   *   * 

“The clear statement doctrine requires that Congress speak with unmistakable clarity before federal preemption of ‘a decision of the most fundamental sort for a sovereign entity’ will be considered. The rule is in ‘acknowledgment that the States retain substantial sovereign powers under our constitutional scheme, powers with which Congress does not readily interfere.’ In Gregory v. Ashcroft (1991), the Court reiterated its longstanding jurisprudential requirement that “[I]f Congress intends to alter the ‘usual constitutional balance between the States and the Federal Government,’ it must make its intention to do so ‘unmistakably clear in the language of the statute,’” and that “Congress should make its intention ‘clear and manifest’ if it intends to pre-empt the historic powers of the States…. No fair reading of Section 706 can find any clear statement of congressional intent that the FCC can interpose itself between states and their political subdivisions. And Section 706 cannot be read to clearly state that Congress intended to preempt state authority over decisions about whether and to what extent to allow its political subdivisions to offer proprietary services.”

*   *   *

“Finally, and importantly, the FCC’s preemption of state restrictions on government-owned broadband networks violates constitutional federalism principles. The Supreme Court has stressed that: ‘The Framers explicitly chose a Constitution that confers upon Congress the power to regulate individuals, not States.’ The Constitution established ‘two orders of government, each with its own direct relationship, its own privity, its own set of mutual rights and obligations to the people who sustain it and are governed by it.’ Indeed, ‘[t]he Constitution thus contemplates that a State’s government will represent and remain accountable to its own citizens.’

Local governments are created by state constitutions through state legislation. They are accountable to the citizens of the respective states in which they exist. Thus, the Supreme Court has long recognized that “[s]tate political subdivisions are ‘merely ... department[s] of the State, and the State may withhold, grant, or withdraw powers and privileges as it sees fit.’” Our constitutional regime does not recognize, as a matter of legal status, ‘citizens’ of Chattanooga or Wilson. It does recognize citizens of Tennessee and North Carolina. And the Constitution confers upon these citizens of states the authority to exert their will through their elected representatives to adopt laws that restrict municipal activities. In essence, this is what the Supreme Court reaffirmed in Nixon, declaring that ‘preemption would come only by interposing federal authority between a State and its municipal subdivisions, which our precedents teach, are created as convenient agencies for exercising such of the governmental powers of the State as may be entrusted to them in its absolute discretion.’”

Internet Political Speech Should Be Reclaimed from Regulation, Not Disclaimed

The Federal Election Commission (FEC) is receiving public comments through November 14 on aspects of its Internet speech regulations. The FEC's request for comments stems from recent requests by Google and Facebook for exemptions from federal campaign speech disclaimer regulations for limited-character online text ads. Specifically, the FEC is deciding whether and how it should revise – and possibly make exceptions to – its regulations regarding disclaimers on online advertisements, such as banners, popups, and short text ads.

The short answer to the FEC is that it should, at minimum, adopt an exception for character-limited ads. If anything should qualify for "small items or impracticable" exceptions, it's online ads containing less than 160 characters. But the FEC's prospective action also provides occasion to reflect more generally on the regulation of political speech on the Internet. Do we need federal regulation of political speech on social networks, blogs, and other Web sites? The interactive and information-rich nature of the Internet as well as the fact that such disclaimer statements directly burden online speech suggests that the federal government should consider retreating from regulating Internet communications.

As explained in its Advance Notice of Proposed Rulemaking, published in the Federal Register, the Bipartisan Campaign Reform Act of 2002 ("BCRA") expanded and made more specific disclaimer requirements for certain "public communications." Disclaimers are required for public communications that: (1) are made by a political committee; (2) expressly advocate election or defeat of a clearly identified federal candidate; or (3) solicit a contribution. Under BCRA and FEC regulations, a "disclaimer" statement must appear on such communications, identifying who paid for it and whether a candidate authorized it.

The FEC "adopted its current rules governing Internet communications in 2006 in response to the decision of the U.S. District Court for the District of Columbia in Shays v. FEC [2004] ("Shays I"). There the District Court ruled, among other things, that "[w]hile all Internet communications do not fall within [the scope of 'any other form of general public political advertising'], some clearly do." The District Court left it to the Commission to determine "what constitutes 'general public political advertising' in the world of the Internet," and thus should be treated as a "public communication."

Disagreement among FEC Commissioners resulted in the FEC not appealing the District Court's ruling regarding the FEC's prior exclusion of all Internet communications from the definition of "public communication." Instead, following Shays I, the FEC added "Internet communications placed on another person's Web site for a fee" to the regulatory definition of "public communication." This definition swept in all potential forms of Internet advertising, such as banners, streaming videos, popups, and directed search results. And by including Internet communications placed for a fee on another person's Web site in the definition of "public communication," those Internet communications became subject to BCRA's disclaimer requirement.

As mentioned earlier, Google and Facebook recently requested exemptions from FEC regulations relating to disclaimers for short-character text ads. Google asked the FEC if it could sell search engine-generated text ads of around 95 characters to candidates and political committees if those ads didn't include disclaimers. The FEC issued an Advisory Opinion, concluding such ads were not in violation of BCRA, although the Commissioners could not reach agreement on the grounds of their ruling. Facebook subsequently asked the FEC if its ads limited to 160 characters qualified for either the "small items" or "impracticable" exceptions to the disclaimer requirements. But FEC Commissioners could not reach agreement and never issued an advisory opinion in response to Facebook's request.

That the FEC could not even agree to exempt Facebook ads and that its Notice suggests imposing alternative disclaimer requirements on small online ads through the use of required hyperlinks, micro bars, or buttons, suggests the agency might favor a pro-regulatory approach for dealing with online political ads in the future. At this point, a FEC general exemption for character-limited online ads is no sure thing.

Hopefully, the FEC will recognize that disclaimer requirements are especially burdensome when it comes to online advertising along the lines of Facebook updates, Tweets, and other social networking services. Space and character length come at a premium in such cases, increasing the likelihood that disclaimer messages will interfere with ad purchaser messages. To the extent that disclaimer restrictions make innovative kinds of online advertising less attractive to ad buyers, it can also lessen the ability of online content companies to rely on ad sales to monetize their services and be able to offer them to the public for free. After all, typical online services from Google, Facebook, Twitter, and the like are ad-supported and cost nothing for regular consumers to sign up for and use.

Moreover, nothing in Shays I expressly or impliedly requires limited character ads on the Internet to be swept into the FEC's regulatory sweep. So the FEC has reason enough to grant an exemption.

Yet for those who appreciate the transformative effect of the Internet on media and speech communications, the FEC's dabbling with online ads and disclaimer requirements brings to the surface broader questions about whether the federal government should even be regulating Internet communications.

For starters, a regulatory approach to online political speech fails to appreciate the Internet's low barriers to entry. Users can offer ideas and counter other ideas without having to own a complex media empire, let alone a broadcast station, channel, or program. Blogs, streaming video such as YouTube, and social networking services have made increasingly simple and inexpensive the avenues for public participation in the marketplace of ideas, political debates, and political elections.

Federal regulation of online political speech also has a paternalistic cast to it, as if Internet users are too easily taken in by online ads advocating a particular political viewpoint or somehow insufficiently able to exercise independent judgment, seek verification, or consult contrary viewpoints. Internet users have near total control over the kinds of content they wish to access or receive. And Internet users enjoy the benefit of easy access to further sources of information concerning the contents of online ads, political or otherwise.

Much more could be said regarding regulation of political speech online – or regulation of political speech made through all other modern media platforms, for that matter. But disclaimer requirements for political speech in modern media platforms, such as the Internet, burden speakers' ability to convey their own message. In particular, this aspect of disclaimer requirements renders them more burdensome than filing periodic reports with the FEC pursuant to campaign finance disclosure requirements.

In the years since BCRA and Shays I, legislation has percolated through Congress – such as the proposed Online Freedom of Speech Act – that would exclude all Internet communications from BCRA's definition of "public communication" and hence from its disclaimer requirements. The idea of subjecting short-character Internet ads to disclaimer requirements, now before the FEC, should cause the Congress and the general public to be concerned about the whole idea of regulating Internet communications.