The AI Election Ad Surge: Why the 2026 Midterms Need a Common Transparency Baseline

The 2026 midterms are becoming a stress test for political-ad transparency. 

Industry forecasts now place total cycle spending between $10.4 billion and $10.8 billion—the most expensive non-presidential election on record—with connected TV alone expected to attract roughly $2.5 billion. Campaigns are not merely buying more media; they are producing more versions, for more screens, at unprecedented speed. 

Generative AI is central to that acceleration. Campaigns can use it to draft scripts, translate messages, generate images and voiceovers, clone approved creative into hundreds of audience-specific variations and optimize delivery. Those efficiencies also increase the chance that synthetic depictions will be mistaken for real speech or events—and that sponsor, provenance and disclosure data will fall away as creative moves among campaigns, agencies, model providers, platforms and publishers. 

Federal rules provide only part of the answer. The Federal Election Commission’s internet-ad disclaimer rules identify payors and authorization status, but the agency declined in 2024 to create a dedicated AI-ad rule, instead confirming that existing fraudulent-misrepresentation rules are technology neutral. That leaves no uniform federal requirement telling voters when political content was materially generated or altered by AI. 

States have moved quickly into the gap. By June 2026, 31 states had enacted laws regulating election deepfakes. Twenty-eight generally require disclosures on synthetic political media; Minnesota and Texas prohibit certain political deepfakes during defined pre-election periods; Maryland applies a year-round prohibition; and Colorado and Utah add metadata requirements. Maine joined the field in 2026 with a law requiring disclaimers on materially manipulated or AI-generated campaign communications. The complex regulatory landscape continues to evolve. 

These laws sit atop already demanding campaign-ad regimes. California uses sponsor- and format-specific disclosures, including separate rules for certain online-platform ads. Washington requires sponsor identification and detailed commercial-advertiser records concerning purchasers, costs, targeting, and distribution. Definitions, timing windows, label wording, remedies, exceptions and technical specifications still vary—and constitutional challenges, including a 2025 ruling striking down part of California’s deepfake framework, add uncertainty. 

For national campaigns and ad-tech companies, compliance is therefore not a single checklist. It changes by state, medium, sponsor, creative technique, audience and proximity to Election Day. A direct disclosure suitable for a television spot may not work in a six-second video, an audio ad or a small mobile placement. A platform’s voluntary AI label may not satisfy a state’s wording, duration, metadata, archive or recordkeeping rules. 

This is why self-regulation is more important than ever—not as a substitute for law, but as a complementary baseline. The Digital Advertising Alliance’s Political Advertising Principles can help companies identify the common core across jurisdictions: sponsor identity, accessible notice, material disclosure, provenance, durable records and accountability. The purple Political Ad icon can provide an interactive route to verified sponsor and campaign information while coexisting with any direct notice a state requires. 

BBB National Programs’ Digital Advertising Accountability Program enforces those principles, educates campaigns and ad-tech participants, monitors compliance and may refer appropriate matters to government agencies. Its guidance can translate shared principles into day-to-day controls while legal teams apply stricter state overlays. 

Three practices should become standard: 

1. Label AI at Creation and Preserve Provenance 

Campaigns should record material generation or alteration when creative is made and carry that information—along with sponsor, version, authorization and targeting data—through the supply chain. 

2. Use a Universal Floor With State-Specific Overlays 

Every ad should provide reliable sponsor information, perceptible AI notice where material, an interactive Political Ad mechanism and retrievable records. Rules engines can then add required language, timing, font, audio, metadata, repository and reporting elements by jurisdiction and format. 

3. Test Disclosures Before Launch

Teams should verify readability, audibility, persistence, destination accuracy and record retention across CTV, social video, display, audio, mobile and programmatic inventory—and maintain a rapid process to correct or withdraw defective creative. 

The 2026 midterms will test whether political-ad infrastructure can keep pace with record spending, synthetic media and rapidly changing state obligations. Law sets indispensable boundaries, but it cannot by itself create one usable workflow across dozens of regimes. A strong self-regulatory baseline can supply that connective tissue: interoperable, auditable practices that respect stricter state protections, while giving voters consistent answers to two essential questions: who is speaking, and how was this message made? 

Dr. Divya Sridhar is the Vice President, Global Privacy Initiatives & Operations at BBB National Programs. Her portfolio encompasses the design, development, and launch of multiple industry self-regulation programs, including the Global Privacy Division’s Cross Border Privacy Rules (CBPR) program and the EU-U.S. Data Privacy Framework Services program, the Digital Advertising Accountability Program, the Digital Health Privacy Program and the TeenAge Privacy Program. Sridhar is a seasoned leader focused on data privacy and emerging technology policies at the international, federal, and state level. In previous roles, Divya served in numerous capacities at think tanks, private companies, and nonprofits leading government affairs and policy work, specializing across the health and education technology sectors. She has written books and authored publications in the fields of tech, privacy and public policy.