Election Law - Influencing a Voter By Dissemination of a Deepfake - Prohibition
HB525 amends Maryland election law to prohibit using fraud to influence or attempt to influence a voter’s decision, and it specifically adds the use of synthetic media and deepfakes to the law’s definition of fraud. The bill creates a new offense for knowingly and fraudulently disseminating a deepfake to affect whether a voter casts a ballot, or how they vote on a candidate or ballot issue. It also defines key terms such as “influence,” “synthetic media,” and “deepfake,” with the latter covering realistic but false images, audio, or video created or manipulated using generative artificial intelligence or other digital technology.
The bill also preserves exceptions for deepfakes disseminated by the news media and for satire or parody. It further provides that certain intermediaries and service providers—such as internet platforms, cloud providers, broadcasters, cable and streaming operators, and newspapers or websites—may not be held liable solely because their services are used by another person to disseminate a prohibited deepfake. Violations are misdemeanors punishable by a fine of up to $5,000, imprisonment for up to 5 years, or both, and the act takes effect June 1, 2025.
HB525 expands Maryland’s election-fraud and voter-intimidation provisions in Article 16 of the Election Law by expressly covering AI-generated or digitally manipulated synthetic media used to deceive voters. It adds a new deepfake-specific prohibition in § 16-905 and amends § 16-201 to include synthetic media within the meaning of fraud, thereby broadening the conduct that can trigger criminal penalties in election-related deception cases. The bill affects candidates, campaigns, political actors, media entities, online platforms, and other communications providers, while carving out protections for news reporting, satire, parody, and certain service providers.
The bill appears to have broad legislative support, passing third reading with 126 yeas and 10 nays. That vote suggests a generally favorable view of the measure as a response to election misinformation and AI-enabled deception. The absence of committee transcript material limits insight into detailed debate, but the strong floor vote indicates overall consensus on the need to address deepfakes in elections.
The main policy tension in HB525 is between preventing deceptive election interference and avoiding overreach into protected speech or intermediary liability. The bill addresses that concern by exempting news media, satire, and parody, and by shielding internet, broadcasting, cloud, and publication services from liability when others misuse their services. Any opposition likely centered on the breadth of the deepfake definition, the criminal penalties, and the risk that the law could chill political speech or legitimate media use, while supporters likely emphasized protecting voters from AI-driven fraud and manipulation.