Pennsylvania AI Deceptive Advertising Law

Overview

The Artificial Intelligence Deceptive Advertising Act was enacted in Pennsylvania as Act No. 99 of 2026 on October 22, 2026, following passage of Senate Bill 806. Section 1 establishes the short title, section 2 confines the subject to deceptive advertising, and section 6 provides commencement 60 days after approval. The effective date is December 21, 2026.

The central rule addresses advertisements using AI-generated or significantly AI-modified content to create a false consumer perception about the nature, origin or authenticity of advertised goods or services. It is a targeted disclosure obligation, not a general prohibition on AI advertising or a labeling requirement for every AI-assisted advertisement. Section 4 combines a conduct standard, a deception trigger, disclosure requirements and exceptions. Section 5 provides enforcement by the Attorney General and appropriate district attorneys, with judicial remedies. An AI disclosure does not extinguish other legal recourse.

Definitions

Section 3 defines artificial intelligence as a machine-based system that, for human-defined objectives, can make predictions, recommendations or decisions influencing real or virtual environments. Its capabilities include perceiving those environments, automatically abstracting perceptions into models through analysis, and using model inference to formulate information or action options. The definition includes generative AI: models that emulate the structure and characteristics of input data to generate synthetic content. Synthetic content means information, including an image, video, audio clip or text, significantly modified or generated by AI.

Clear and conspicuous disclosure is a cumulative, presentation-sensitive standard. Size, color, contrast, location, duration and audibility must make the disclosure readily noticeable, readable, understandable and capable of being heard. It must use the language or other interaction through which the consumer interacts with or comprehends the synthetic content, and must not contradict accompanying information. Explanatory or clarifying disclosures must be nearby and unobscured. Audio requires sufficient volume and cadence; visual disclosure requires sufficient size, shade and screen duration. Print advertisements and promotional materials, including point-of-sale displays and brochures, require noticeable type size, placement and background contrast.

Governance and Institutional Framework

Section 5 assigns enforcement to existing public legal authorities. The Pennsylvania Attorney General or appropriate district attorney may seek an injunction when there is reason to believe a person is using, or is about to use, a method, act or practice declared unlawful by section 4, and proceedings would be in the public interest. The action is brought in the Commonwealth’s name and may request temporary or permanent restraint. This permits intervention against anticipated prohibited conduct as well as conduct already occurring, subject to the statutory conditions.

The Attorney General or appropriate district attorney may also bring a civil action for violations. A court of competent jurisdiction determines judicial relief under section 5. The core disclosure obligation addresses an advertiser’s creation and dissemination of covered advertisements, including causing those activities. Responsibility turns on conduct under section 4 rather than participation in the AI supply chain alone. Technology developers and advertising platforms are not assigned universal obligations simply because they supply technology or distribution services. The institutional framework is public enforcement through the courts rather than AI licensing, certification, registration or a specialist appeals tribunal.

Disclosure Trigger and Exceptions

Section 4(a) prohibits an advertiser from knowingly, recklessly or willfully creating and disseminating, or causing the creation and dissemination of, an advertisement using synthetic content in a manner creating a false consumer perception about the nature, origin or authenticity of advertised goods or services without clear and conspicuous disclosure. The disclosure must state that the advertisement contains content modified or generated using artificial intelligence. The conduct standard and specified false perception are both central to the trigger; AI involvement somewhere in an advertising workflow is not by itself sufficient.

Section 4(b)(1) excludes content evidently fictional based on its content or the context of the depiction. Section 4(b)(2) excludes parody, satire, commentary and criticism. Section 4(b)(3) excludes content limited to text. Although text is included in the definition of synthetic content, that inclusion does not override the text-only exception. An exception removes this disclosure obligation rather than authorizing conduct prohibited by other law. Section 4(c) expressly preserves any recourse under law, so an AI label does not provide general immunity for deceptive claims or other unlawful advertising.

Implementation Framework

For covered advertisements, section 4(a)(1) requires disclosure in the first instance whenever synthetic content is presented to the consumer. Section 4(a)(2) requires disclosure in the same medium as the synthetic content. Section 4(a)(3) requires the disclosure, to the extent technically feasible, to be permanent or extraordinarily difficult to remove from the synthetic content. Technical feasibility qualifies permanence and removal resistance, rather than the first-presentation or same-medium duties. The functional standard allows implementation suited to the advertisement without prescribing a particular watermarking technology or labeling platform.

A practical compliance review can examine production history, whether an asset was generated or significantly modified by AI, the consumer-facing meaning of the advertisement, applicable exceptions, and knowing, reckless or willful conduct. For a covered advertisement, implementation must address disclosure substance, timing, medium, noticeability, comprehension, consistency and proximity. Audio requires attention to volume and cadence; visual and print presentation require appropriate size, contrast, placement and duration where applicable. Internal reviews are practical recommendations, not separate statutory procedures. The statutory criteria are functional rather than fixed font sizes, minimum display seconds, approved wording templates or prescribed technical standards.

Monitoring and Evaluation

The Act’s monitoring structure is enforcement-oriented. Public authorities may act on reason to believe unlawful practices are occurring or are about to occur, with public-interest conditions applying to injunction proceedings. Advertisers remain responsible for satisfying section 4 when covered advertisements are created and disseminated. The statutory framework does not establish periodic regulatory reporting, mandatory audits, incident notifications, an evaluation schedule, annual assessments or publication of compliance metrics.

Voluntary quality checks can assess disclosures in the form consumers actually encounter. Editing, cropping, compression, translation and distribution changes can obscure a visual label, impair an audio statement or separate explanatory information from its disclosure. Reviewing those changes is a practical application of the presentation requirements rather than an independently mandated testing protocol. Maintaining asset versions, disclosure examples and technical-feasibility assessments can support internal review. Such documentation is voluntary rather than subject to an Act-specific retention period, and does not itself create an automatic compliance presumption through documentation, audit results or use of a technical tool.

Penalties, Liability, and Judicial Relief

Section 5(a) authorizes temporary or permanent injunctions under its reason-to-believe and public-interest conditions. Section 5(b) authorizes civil actions by the Attorney General or appropriate district attorney. Under section 5(c), a court of competent jurisdiction may award damages for a violation and may award a civil fine not exceeding $3,000, together with actual attorney fees and court costs. The $3,000 amount is the ceiling for the specified civil fine, not a cap on every form of monetary relief. The Act supplies a civil enforcement framework rather than a criminal offense, enhanced repeat-offender schedule or mandatory minimum fine.

Section 4(c) preserves any recourse under law despite disclosure of AI use. An advertiser must therefore assess underlying claims separately from whether an AI label is present. Section 5 identifies public officials as civil enforcement actors; independently available legal remedies remain preserved by section 4(c). The Act does not expressly create a consumer private cause of action or specify a fine-calculation unit such as advertisement, impression, consumer or day. Judicial relief is provided without an Act-specific administrative reconsideration process, cure period, special appeals procedure or appeal deadline.

Relationship to Other Instruments

The Act contains its own short title, definitions, disclosure rule, enforcement provisions and commencement clause. It establishes a targeted AI advertising disclosure regime while preserving other legal recourse under section 4(c), rather than replacing existing advertising protections. Its operative structure is set out in the final legislative text, Printer’s No. 1808. The Act does not identify amendments to named provisions of another statute or incorporate an external technical standard, certification framework or model-risk methodology.

The subject is deceptive advertising, rather than training-data regulation, rights over a person’s likeness, mandatory model evaluations or a general privacy regime. Those issues can arise in an advertising project, but are distinct from the duties established here. An advertiser should assess the underlying advertising claims independently of the presence, absence or technical quality of an AI disclosure. Likewise, fictional, expressive or text-only content falling within section 4(b) remains subject to other applicable legal restrictions because the exception concerns this particular disclosure obligation.

State and Federal Scope

This is a Pennsylvania state enactment, not a federal AI law or nationwide advertising regulation. Enforcement is assigned to Pennsylvania’s Attorney General and appropriate district attorneys, and restraint actions are brought in the Commonwealth’s name. The substantive scope remains deceptive advertising despite the AI terminology in the title and definitions. It does not establish a general regime for telecommunications networks, AI model development or all synthetic media distributed throughout the United States.

The Act’s compliance pathway is its own disclosure standard and state judicial enforcement. It does not delegate authority to a federal agency, incorporate a federal AI disclosure standard, establish federal certification as a compliance pathway, or provide a federal-compliance safe harbor. The preservation of any recourse under law means a Pennsylvania disclosure is not an exhaustive resolution of an advertisement’s legality. Wider federal advertising obligations remain a separate consideration; the implementation deadline established by section 6 is the Pennsylvania effective date of December 21, 2026.

Implementation Timeline

The legislative milestones are recorded in the official Senate Bill 806 information and legislative history. The bill progressed from Printer’s No. 0871 to amended Printer’s No. 1690 and then Printer’s No. 1808 before enactment.

Sources and References