Political advertising can reach publisher inventory through the same programmatic demand channels as any other campaign. For publishers, however, serving political demand comes with a different set of considerations. Political advertising is regulated differently across markets, and the rules that matter are no longer limited to the candidates, campaigns, and political organizations paying for the ads.
Many election laws still place their obligations on advertisers. Beginning in early 2025, a newer set of global regulations reach further into the advertising supply chain, introducing transparency, recordkeeping, and other requirements for publishers and intermediaries involved in serving political ads.That shift puts a portion of the compliance burden directly on the publisher running the ad, regardless of the channel, not only on the campaign that paid for it. A growing patchwork of AI and synthetic media laws adds another consideration, with disclosure requirements tied to how political creative is generated or altered and responsibilities that vary by jurisdiction.
For publishers, political ad compliance in 2026 is therefore not simply a question of whether political advertising is permitted. It also depends on what is being served, where it appears, and what obligations apply to the parties involved in its delivery.
Political advertising illustrates this challenge in its own way. Keeping up with the rules isn’t about understanding one law. It’s about navigating a patchwork of regional regulations, platform policies, and fast-moving AI disclosure requirements across the markets where publisher inventory reaches an audience. What follows is a reference point for understanding that landscape, not a substitute for legal advice specific to where an ad is served.
What Counts as Political Advertising?
Political ads are not always easy to identify at the point of delivery. A direct candidate ad may be obvious, while an issue ad, referendum campaign, or creative built around a political message can be harder to classify consistently. Part of the problem is that there is no universal legal definition of political advertising.
The EU’s Transparency and Targeting of Political Advertising Regulation (TTPA), Regulation (EU) 2024/900, takes a broad approach. It covers advertising by, for, or on behalf of a political actor. It also covers advertising that is liable and designed to influence the outcome of an election or referendum, voting behavior, or a legislative or regulatory process. That can bring paid issue-based advertising within scope even when no candidate is named. Political views expressed in a personal capacity are excluded, as is editorial content under editorial responsibility unless its preparation or publication involves specific third-party payment or remuneration.
In the United States, the Federal Election Campaign Act’s disclaimer rules for express advocacy use a narrower, candidate-focused standard, covering communications that expressly advocate the election or defeat of a clearly identified candidate. Other federal rules cover certain election-related communications. A creative that falls within the TTPA’s definition may therefore not meet the same threshold under a US federal rule.
Recognition is the actual compliance problem for publishers. The right controls can only be applied once political content has been correctly identified, but programmatic categories do not necessarily map neatly to these legal definitions, and what counts as political advertising can depend on the creative’s message, sponsor, purpose, and where it is served.
The EU TTPA: What Publishers Need to Know
The Transparency and Targeting of Political Advertising Regulation (TTPA), Regulation (EU) 2024/900, creates EU-wide rules for the transparency and targeting of political advertising. It was adopted on March 13, 2024, entered into force on April 9, 2024, and most of its provisions became applicable on October 10, 2025. Unlike many election laws, it regulates more than the campaigns and political organizations buying ads, creating obligations across the political advertising chain, including for a defined category of “political advertising publisher.”
Where the TTPA applies
The TTPA covers political advertising published or disseminated in the EU, made public in one or more member states, or directed to individuals in one or more member states, regardless of where the service provider or sponsor is established. A publisher therefore does not necessarily fall outside the regulation simply because the business itself is based outside Europe. Non-EU providers offering covered political advertising services in the Union may also be required to designate a legal representative in an applicable member state.
What publishers have to do
A political advertising publisher, broadly a provider that publishes, delivers, or disseminates political advertising, has specific transparency duties. Political ads must be labelled as such and accompanied by an accessible transparency notice containing information about the sponsor, funding, relevant political process, and targeting where applicable. Publishers also have recordkeeping and reporting obligations, with required information generally retained for seven years.
Online political ads must also be made available through the European political advertising repository. VLOPs and VLOSEs must make ads and their transparency information available from publication, while other online political advertising publishers have up to 72 hours after first publication. Commission Implementing Regulation (EU) 2026/818, adopted in April 2026, sets the technical arrangements for the repository. As of mid-2026, however, the schema definitions, API endpoints, and integration documentation needed to actually submit to it had not yet been published, and no go-live date had been announced, meaning the legal obligation currently exists ahead of the mechanism to comply with it.
What happens if a publisher doesn’t comply
Member states set and enforce penalties, so there is no single fixed fine across the EU. The TTPA nevertheless sets a significant ceiling: for relevant infringements, maximum financial penalties must be capable of reaching 6% of annual income or budget, or 6% of worldwide annual turnover in the preceding financial year.
How Google and Meta responded
The compliance burden has already affected political advertising in the EU. Google stopped serving political ads across its EU platforms, including YouTube, in late September 2025, ahead of the TTPA becoming generally applicable on October 10. Meta followed, ending political, electoral, and social issue ads across Facebook and Instagram in the EU on October 6. Both pointed to the complexity and legal uncertainty created by the new regulatory environment.
The TTPA does not ban political advertising, and neither company was required to exit the category. Their decisions instead illustrate the operational question facing publishers that choose to accept political demand: whether its value justifies the additional compliance requirements.
AI Political Ad Disclosure Laws Are Creating a New Compliance Layer
AI-generated and digitally altered political content is increasingly subject to rules of its own, layered on top of existing political advertising requirements. The trend is particularly pronounced in the US, but unlike the TTPA, these laws do not follow a common model for publisher responsibility. In many cases, the disclosure duty sits primarily with the campaign or advertiser creating the content, while liability elsewhere in the distribution chain varies by jurisdiction.
The US state-by-state patchwork
By mid-2026, around 30 states had enacted laws addressing election deepfakes or synthetic political media. The requirements vary in the content they cover, how close to an election they apply, what disclosures are required, and who can be held responsible.
Louisiana’s HB 459, signed as Act 806 on June 9, 2026, and effective August 1, 2026, requires a clear disclosure on covered electioneering communications that portray a candidate using an AI-created image or likeness with no window tied to proximity to an election. Violations can carry fines of up to $2,000, up to two years’ imprisonment, or both, and liability attaches where the intent is to injure a candidate’s reputation or to deceive a voter, and the law exempts media entities that had no input into or control over the content they carried.
Vermont’s Act 75 (S.23), signed March 5, 2026, takes a different approach. It requires disclosure on deceptive synthetic media distributed within 90 days of an election, with penalties that scale by circumstance: up to $1,000 as a baseline, rising to $5,000 where there is intent to cause violence or bodily harm, $10,000 for a repeat violation within five years, or $15,000 where both apply. It also exempts certain broadcasters, bona fide news programs and publications that flag the content’s authenticity, telecommunications and information-service providers, providers of interactive computer services acting on content supplied by another person, and satire or parody.
Both laws illustrate an important distinction for publishers: an advertiser’s obligation to disclose AI-generated political content does not automatically make every party involved in distributing the ad equally liable. Identifying an ad as AI-generated political content is therefore only the first step. Whether an obligation follows depends on the state, the type of content, and the publisher’s role in its distribution.
Brazil
Brazil’s 2026 election rules reach more directly into the advertising and application layer. The Superior Electoral Court approved Resolução nº 23.755/2026 on March 2, 2026, published it on March 4, and it took effect immediately, amending the country’s existing electoral advertising framework, Resolução nº 23.610/2019, for the 2026 elections.
The rules require disclosure of AI-generated or manipulated electoral content, and applications offering paid political-content boosting must provide advertisers with a field to declare the use of AI or equivalent technologies. They also prohibit publishing, republishing, or paid boosting of certain synthetic content involving candidates or public figures from 72 hours before voting through 24 hours after.
For online services covered by these requirements, the clearly established consequence of non-compliance is immediate removal of the content or unavailability of the communication service itself, either proactively or by judicial order. Unlike many US state disclosure laws, Brazil’s framework does not stop at what the campaign must disclose: parts of the compliance obligation extend directly to the services used to distribute and promote political content.
Political Ad Rules Publishers Should Know in Other Markets
Political advertising rules in other major markets take different approaches to the distribution layer. Some impose obligations directly on qualifying online platforms, while others primarily regulate the campaign or advertiser responsible for the creative.
United States: Federal disclaimer requirements
At the federal level, the Federal Election Campaign Act (FECA), 52 U.S.C. §30120, requires covered political communications to identify who paid for them and, where applicable, whether they were authorized by a candidate or candidate committee. FEC regulations at 11 CFR §110.11 extend this disclaimer requirement to paid internet communications placed on another person’s website, digital device, application, or advertising platform. The obligation follows the person paying to place the communication rather than automatically making the publisher serving it responsible, making this an important baseline for what compliant US federal political creative should contain.
Canada
The Canada Elections Act requires qualifying online platforms that sell advertising space directly or indirectly to maintain a registry of regulated political advertising during federal election periods. The requirement applies above Canadian traffic thresholds of 3 million average monthly visitors for primarily English-language platforms, 1 million for French-language platforms, and 100,000 for platforms primarily in another language. The framework was introduced through the Elections Modernization Act, which received Royal Assent on December 13, 2018, with the relevant provisions coming into force in 2019. Canada also prohibits selling advertising space to a foreign person or entity for the purpose of transmitting an election advertising message. Failure to maintain the required registry is an offence under the Act, and the Commissioner of Canada Elections can respond with measures including an administrative monetary penalty, a compliance agreement, or prosecution.
India
India uses a pre-publication model for political advertising during elections. The Election Commission of India (ECI) requires political parties, candidates, and other covered advertisers to obtain pre-certification from a Media Certification and Monitoring Committee (MCMC) before releasing political ads through covered electronic and digital media. Rather than stemming from one newly enacted law, the requirement is applied through ECI directions for individual election cycles; it was reiterated on March 20, 2026, for the Assembly elections in Assam, Kerala, Puducherry, Tamil Nadu, and West Bengal. The certification duty sits with the political advertiser seeking to release the ad rather than creating a general certification duty for the web publisher carrying it, and enforcement focuses primarily on the party or candidate responsible through the ECI’s electoral enforcement powers.
United Kingdom
The UK’s digital imprint regime was introduced by the Elections Act 2022 and came into force on November 1, 2023. Paid digital material meeting the definition of political material must include an imprint identifying the promoter and anyone on whose behalf it is being published, across formats including websites and digital advertising. The legal responsibility sits with the promoter and anyone on whose behalf the material is published, rather than automatically with the website serving the creative. Publishing covered material without the required imprint is an offence under Section 48 of the Act. The Electoral Commission can require non-compliant material to be removed, while a court can also order removal following a conviction.
Australia
At the federal level, the Commonwealth Electoral Act 1918 requires certain electoral communications, including paid online electoral advertising, to carry an authorisation identifying the person or entity responsible and the required address details. The authorisation regime in Part XXA has applied since March 2018 and covers electoral matter communicated to people in Australia. As in the UK, carrying the ad does not automatically transfer the advertiser’s legal responsibility to every publisher or ad-tech intermediary involved in serving it. The Australian Electoral Commission monitors compliance through complaints and its own reviews and can respond to suspected breaches through measures including education, warning correspondence, and further enforcement action where appropriate.
Why Political Ads Are an Ad Quality Problem, Not Just a Legal One
Political advertising is not only a compliance issue. Political creative can be miscategorized or disguised as it moves through programmatic supply, and the signals attached to an ad do not necessarily tell publishers what the creative actually contains. An ad can meet the compliance requirements that apply to it and still be something a publisher would not have chosen to run.
AI-generated content adds another layer. The same synthetic media used in legitimate political campaigns can also be used to impersonate candidates, elected officials, or other recognizable figures in scams and deceptive advertising. In those cases, political imagery or messaging is simply the vehicle for a different ad-quality problem. The presence of a political figure does not necessarily mean the advertiser is a political campaign, or even that the ad is genuinely political.
Brand suitability creates a separate concern. A political ad can be authentic, legitimate, and legally compliant while still associating a publisher’s inventory with a candidate, issue, or message it would rather not carry. In programmatic environments, that makes political advertising a category that requires active oversight rather than a simple compliant-or-non-compliant decision.
How Publishers Can Control Political Advertising
Platform-level category settings, demand-source controls, and allow or block lists are a useful starting point, but they depend on how an ad has been classified upstream, and that classification is not always accurate, accidentally or intentionally. Advertisers declare their own campaign categories in parts of the programmatic ecosystem, which gives a motivated advertiser a way to route political content past category-based controls entirely. A category label indicates how a creative has been classified; it does not necessarily tell a publisher what the creative actually contains.
That is where GeoEdge adds another layer of control. GeoEdge analyzes creative imagery and text, as well as landing pages, independently of the advertiser’s declared category. Publishers can define their own policies for what should and should not appear on their sites, while GeoEdge identifies and blocks unwanted ads in real time. For political advertising, this provides granular control based on the content actually being served rather than relying solely on how the demand was categorized upstream.
None of this replaces legal judgment. The laws covered throughout this guide determine which obligations apply, while each publisher still has to decide what political advertising it is willing to carry. GeoEdge provides the visibility and control needed to put that decision into practice. Learn how creative-level monitoring and real-time blocking can help publishers take greater control over political advertising in their inventory.
A note on accuracy: Political advertising law is changing quickly, and jurisdictions are actively amending, proposing, and reversing rules covered in this guide. This guide reflects the regulatory landscape at the time of publication and should not be treated as legal advice.
