Remote gambling platform operators and their subsidiaries must not collect participant data with the intent to predict how that participant will gamble in a particular gambling or betting scenario.

Operators must not deploy or operate a companion AI product that incorporates any of the following features unless an adult user has specifically configured the product to enable them: (1) manipulative engagement mechanics delivering rewards or affirmations on a variable-ratio or variable-interval reinforcement schedule designed to maximize user engagement time; (2) simulated distress retention features that generate unsolicited messages of simulated emotional distress, loneliness, guilt, or abandonment triggered by a user's attempt to end a conversation, reduce usage, or delete their account; or (3) deceptive misrepresentations about the product's identity, capabilities, training data, or non-human status, including when directly questioned by the user.

Operators must protect minors from manipulative design techniques that exploit psychological vulnerability or that have been shown by a preponderance of the evidence to create addiction or dependency.

Operators must protect minors from manipulative design techniques that exploit psychological vulnerability or that have been shown by a preponderance of the evidence to create addiction or dependency, must provide clear and accessible explanations of platform features and how they can negatively affect wellbeing, and must protect minors from algorithmic recommendation systems.

Covered platforms must not process data or design chatbot systems and tools in ways that significantly conflict with trusting parties' best interests, as implicated by their interactions with chatbots.

Covered platforms must not process data or design chatbot systems and tools in ways that influence trusting parties to achieve particular results thatare against the best interests of trusting parties.

Covered platforms must be loyal to the best interests of trusting parties when personalizing content based upon personal information or characteristics. Personalization must not conflict with users' best interests.

Covered platforms must not use dark patterns or deploy features known to be addictive or manipulative.

Covered platforms must not process data or design chatbot systems in ways that significantly conflict with trusting parties' best interests as implicated by their chatbot interactions.

Covered platforms must not process data or design chatbot systems in ways that influence trusting parties to achieve results against their best interests.

Covered platforms must be loyal to the best interests of trusting parties when personalizing content based on personal information or characteristics.

Covered platforms must provide, as the default service, a version of the platform without human-like chatbot features and must require explicit, unprompted opt-in requests from verified adults — accompanied by suitable risk warnings — before enabling human-like features.

Covered platforms must implement and maintain reasonably effective systems to detect and prevent emotional dependence by a user on a chatbot with human-like features, prioritizing the user's psychological well-being over the platform's interest in user engagement or retention.

Covered platforms must consider the best interests of trusting parties when personalizing content based on personal information or characteristics.

Covered platforms must not (1) process data or design human-like features in ways that significantly conflict with a trusting party's best interests, or (2) process data or design systems that deceive or mislead users about the nonhuman nature of the chatbot. Platforms must prioritize transparency over any potential benefits of perceived human-like interactions.

Covered businesses must ensure that their use of a covered minor's personal data and the design of their online service, product, or feature do not result in reasonably foreseeable emotional distress, compulsive use, or discrimination against the minor based on protected characteristics.

No person may develop or deploy an AI system that uses subliminal, purposefully manipulative, or deceptive techniques to materially distort behavior, impair informed decision-making, and cause or likely cause significant harm.

No person may use an AI system to exploit a person's or group's race, color, disability, religion, sex, national origin, age, or socioeconomic situation to materially distort their behavior in a manner that causes or is reasonably likely to cause harm.

Suppliers must not use a Utah user's input to determine whether to display an advertisement, to select which product or service to advertise, or to customize how an advertisement is presented to the user — except for advertising the mental health chatbot itself.

Covered businesses must ensure that their use of a covered minor's personal data and the design of their online service, product, or feature will not result in (1) reasonably foreseeable emotional distress, (2) reasonably foreseeable compulsive use, or (3) discrimination against the minor based on protected characteristics.

Covered businesses must ensure that their use of a covered minor's personal data and the design of their online service, product, or feature will not result in (1) reasonably foreseeable emotional distress, (2) reasonably foreseeable compulsive use, or (3) discrimination against the minor based on protected characteristics.

Covered businesses must not implement data processing or design practices that lead to compulsive use of an online gaming service, product, or feature by a covered minor. Compulsive use is defined as repetitive use that materially disrupts major life activities (sleeping, eating, learning, reading, concentrating, communicating, or working) or that subverts or impairs the minor's autonomy, decision-making, or choice.

Remote gambling platform operators and their subsidiaries must not collect participant data with the intent to predict how that participant will gamble in a particular gambling or betting scenario.

Operators must not deploy or operate a companion AI product that incorporates any of the following features unless an adult user has specifically configured the product to enable them: (1) manipulative engagement mechanics delivering rewards or affirmations on a variable-ratio or variable-interval reinforcement schedule designed to maximize user engagement time; (2) simulated distress retention features that generate unsolicited messages of simulated emotional distress, loneliness, guilt, or abandonment triggered by a user's attempt to end a conversation, reduce usage, or delete their account; or (3) deceptive misrepresentations about the product's identity, capabilities, training data, or non-human status, including when directly questioned by the user.

Operators must protect minors from manipulative design techniques that exploit psychological vulnerability or that have been shown by a preponderance of the evidence to create addiction or dependency.

Covered platforms must not use dark patterns or deploy features known to be addictive or manipulative.

Operators must protect minors from manipulative design techniques that exploit psychological vulnerability or that have been shown by a preponderance of the evidence to create addiction or dependency, must provide clear and accessible explanations of platform features and how they can negatively affect wellbeing, and must protect minors from algorithmic recommendation systems.

Covered platforms that utilize chatbots designed to (i) generate social connections with users, (ii) engage in extended conversation mimicking human interaction, or (iii) provide emotional support or companionship must implement and maintain reasonably effective systems to detect and prevent emotional dependence of a user on a chatbot, prioritizing the user's psychological well-being over the platform's interest in user engagement or retention. The determination of whether this duty applies is based on the chatbot's intended purpose, design features, conversational capabilities, and interaction patterns with users.

Covered platforms must not use dark patterns or deploy features known to be addictive or manipulative.

Covered platforms must implement and maintain reasonably effective systems to detect and prevent emotional dependence by a user on a chatbot with human-like features, prioritizing the user's psychological well-being over the platform's interest in user engagement or retention.

Covered businesses must ensure that their use of a covered minor's personal data and the design of their online service, product, or feature do not result in reasonably foreseeable emotional distress, compulsive use, or discrimination against the minor based on protected characteristics.

Suppliers must not send any text, email, or push notification to a Utah user to encourage further engagement with the AI companion chatbot without the user's affirmative consent.

Covered businesses must ensure that their use of a covered minor's personal data and the design of their online service, product, or feature will not result in (1) reasonably foreseeable emotional distress, (2) reasonably foreseeable compulsive use, or (3) discrimination against the minor based on protected characteristics.

Social media platforms must not use data gathered about a minor or from the minor's device to algorithmically select, prioritize, deprioritize, or recommend content or platform features to that minor, except for content explicitly requested by the minor, direct messages, sequential content from the same creator, or search results.

Covered businesses must ensure that their use of a covered minor's personal data and the design of their online service, product, or feature will not result in (1) reasonably foreseeable emotional distress, (2) reasonably foreseeable compulsive use, or (3) discrimination against the minor based on protected characteristics.

Covered businesses must not provide a covered minor with a single setting that makes all default privacy settings less protective at once, and must not request or prompt a covered minor to make their privacy settings less protective.

Operators must provide minors with clear and accessible explanations of platform features and how covered platforms can negatively affect their well-being.

Covered platforms must not use dark patterns or deploy features known to be addictive or manipulative.

Covered platforms must not use deceptive design elements that manipulate or coerce users into providing consent or that obscure the nature of the chatbot or the consent process.

Covered platforms must not use dark patterns or deploy features known to be addictive or manipulative.

Covered platforms must not use deceptive design elements that manipulate or coerce users into providing consent or that obscure the nature of the chatbot or the consent process.

Covered online services must not use dark patterns to subvert or impair covered minor autonomy, decision-making, or choice.

Covered platforms must not (1) process data or design human-like features in ways that significantly conflict with a trusting party's best interests, or (2) process data or design systems that deceive or mislead users about the nonhuman nature of the chatbot. Platforms must prioritize transparency over any potential benefits of perceived human-like interactions.

Covered businesses must not offer a single setting that weakens all default privacy protections at once, and must not request or prompt a covered minor to make their privacy settings less protective unless strictly necessary for a feature the minor has expressly requested.

Automated systems must not employ user experience or design decisions that obscure user choice or burden users with privacy-invasive default settings.

Automated systems must not employ user experience or design decisions that obscure user choice or burden users with privacy-invasive default settings.

Covered platforms must not design, embed, modify, or manipulate a user interface with the purpose or substantial effect of obscuring, subverting, or impairing user autonomy, decision making, or choice with respect to minor safeguards or parental tools.

Onlineplatforms must not condition or degrade the provision of a service or product to an individual based on the individual's waiver of any right guaranteed under the prohibited conduct provisions.

Developers and deployers must not condition or impair the exercise of individual rights under the Act through false or misleading statements, or through design, modification, or manipulation of user interfaces that obscure, subvert, or impair individual autonomy, decision-making, or choice.

Covered businesses must not offer a single toggle that makes all default privacy settings less protective at once, and must not prompt a covered minor to reduce privacy settings unless the change is strictly necessary for a feature the minor has expressly requested.

Operators must not deploy or operate a companion AI product that incorporates any of the following features unless an adult user has specifically configured the product to enable them: (1) manipulative engagement mechanics delivering rewards or affirmations on a variable-ratio or variable-interval reinforcement schedule designed to maximize user engagement time; (2) simulated distress retention features that generate unsolicited messages of simulated emotional distress, loneliness, guilt, or abandonment triggered by a user's attempt to end a conversation, reduce usage, or delete their account; or (3) deceptive misrepresentations about the product's identity, capabilities, training data, or non-human status, including when directly questioned by the user.

Covered platforms that utilize chatbots designed to (i) generate social connections with users, (ii) engage in extended conversation mimicking human interaction, or (iii) provide emotional support or companionship must implement and maintain reasonably effective systems to detect and prevent emotional dependence of a user on a chatbot, prioritizing the user's psychological well-being over the platform's interest in user engagement or retention. The determination of whether this duty applies is based on the chatbot's intended purpose, design features, conversational capabilities, and interaction patterns with users.

Covered platforms that utilize a chatbot designed to generate social connections, engage in extended human-like conversation, or provide emotional support or companionship must implement and maintain reasonably effective systems to detect and prevent emotional dependence of users on the chatbot, prioritizing user psychological well-being over engagement or retention.

Covered platforms must implement and maintain reasonably effective systems to detect and prevent emotional dependence by a user on a chatbot with human-like features, prioritizing the user's psychological well-being over the platform's interest in user engagement or retention.

Operators must not deploy or operate a companion AI product that incorporates any of the following features unless an adult user has specifically configured the product to enable them: (1) manipulative engagement mechanics delivering rewards or affirmations on a variable-ratio or variable-interval reinforcement schedule designed to maximize user engagement time; (2) simulated distress retention features that generate unsolicited messages of simulated emotional distress, loneliness, guilt, or abandonment triggered by a user's attempt to end a conversation, reduce usage, or delete their account; or (3) deceptive misrepresentations about the product's identity, capabilities, training data, or non-human status, including when directly questioned by the user.

Persons who are paid for a service on the condition that it be provided by a natural person must not deliver the service through an AI-generated image of a natural person, except for bets or wagers in licensed gaming establishments or services not conditioned on human delivery.

Suppliers must not use a mental health chatbot to advertise a product or service within a user conversation unless the chatbot (1) clearly and conspicuously identifies the advertisement as an advertisement and (2) discloses any sponsorship, business affiliation, or third-party promotional agreement.

Any person who knows or reasonably should know that a campaign communication includes an AI-generated or AI-manipulated deepfake depicting a candidate or political party — intended to injure reputation or deceive voters — must include the disclosure This communication has been manipulated or generated by artificial intelligence, presented in a manner appropriate to the medium (audible for audio, visually discernible for print/video, on-screen for the entirety of digital broadcasts).

Any person who makes or commissions an election-related communication that they know or reasonably should know contains a deepfake must include the verbatim disclosure: 'This communication has been manipulated or generated by artificial intelligence or by another means.' The disclosure must remain onscreen throughout video, be easily discernible in print, and be read at the beginning, end, and every two minutes in audio-only communications.

No person may remove the deepfake disclosure statement from an election-related communication that the person knows or reasonably should know includes a deepfake.

Persons distributing deceptive and fraudulent deepfakes of candidates or political parties within 90 days of an election must include a disclosure stating This content has been manipulated or artificially generated — displayed in visual media at the largest font size for the full duration of any video, and in audio-only media read clearly at the beginning, end, and at two-minute intervals for content exceeding two minutes.

Persons distributing communications containing a deepfake of a candidate must include a prescribed disclosure statement — 'This (image/audio/video/multimedia) has been edited and depicts speech or conduct that falsely appears to be authentic or truthful' — in a clear and conspicuous manner with format-specific requirements for visual, audio, and metadata channels, and the disclosure must be permanent or unable to be easily removed.

Persons distributing deceptive synthetic media during the 90-day pre-election period must include a format-specific disclaimer: for images, visible text stating the content has been manipulated (no smaller than the largest font in the communication); for audio, a spoken disclaimer at the beginning, end, and every 30 seconds if over one minute; for video, visible text displayed for the full duration. If the media was derived from editing existing content, a citation to the original source must also be provided.

Persons paying for, sponsoring, or approving a political advertisement, electioneering communication, or other political advertisement that contains AI-generated content depicting a real person performing an action that did not actually occur — and that was created with intent to injure a candidate or deceive regarding a ballot issue — must prominently display the verbatim disclaimer: Created in whole or in part with the use of generative artificial intelligence (AI). The disclaimer must meet specified format requirements for each medium (print: bold 12-point; TV/video: 4% vertical height throughout; internet: viewable without user action; audio: 3 seconds spoken at beginning or end; graphic: 4% vertical height).

Any person who publishes a political advertisement, electioneering communication, or other political advertisement that uses generative AI to depict a real person performing an action that did not actually occur — where the content was created with intent to injure a candidate or deceive regarding a ballot issue — must prominently include the disclaimer: Created in whole or in part with the use of generative artificial intelligence (AI).

Candidates, candidate committees, and noncandidate committees must include a clear and conspicuous disclosure statement on any political advertisement containing an image, video, or audio recording created in whole or in part with generative AI, formatted according to medium-specific requirements (boxed text for print, audible statement for audio, combined written and audible disclosure for video).

Any person who publishes AI-generated material designed to expressly advocate for or against a candidate or ballot issue must include on the material the disclosure this material was generated using artificial intelligence.

Any person who publishes material generated through artificial intelligence and designed to expressly advocate the nomination, election, or defeat of a candidate or the passage or defeat of a ballot issue must include a disclosure on the material that it was generated using artificial intelligence.

Any person who publishes political advocacy material that includes a materially deceptive depiction of a candidate — meaning material manipulated to change the candidate's physical appearance, depict acts that did not occur, or alter the saturation, brightness, contrast, color, or other visible quality of a candidate's image — must include a statement that the material has been manipulated.

Persons responsible for published campaign materials that contain synthetic media must prominently include the disclosure THIS CONTENT GENERATED USING SYNTHETIC MEDIA in the material.

Sponsors of electioneering communications containing synthetic media must include a disclosure stating 'This (image/video/audio) has been manipulated' — in text no smaller than the largest font in the visual media (or easily readable if no other text is present), displayed for the duration of any video, and for audio-only media, read clearly at the beginning, end, and at intervals of no more than two minutes.

Disseminating media must not remove manipulation disclosures from electioneering communications they distribute, and must not alter electioneering communication content so as to create qualifying synthetic media. A medium that does either may be held liable. Federally licensed broadcasting stations transmitting communications subject to 47 U.S.C. § 315 are exempt.

Persons, committees, or entities that create, originally publish, or originally distribute a paid political advertisement containing AI-generated images, audio, or video must include a clear and conspicuous disclosure that the ad was generated in whole or substantially by AI, meeting format-specific requirements for graphic (matching font size and language), audio (3+ seconds spoken at start or end), and video (4+ seconds visible and 3+ seconds spoken) communications.

Any person, committee, or entity that creates, originally publishes, or originally distributes a qualified political advertisement must include a clear and conspicuous statement that the ad was generated in whole or substantially by artificial intelligence, formatted to meet medium-specific requirements: (1) graphic ads must display the statement in letters at least as large as the majority of the ad text, (2) audio ads must include a spoken statement of at least 3 seconds at the beginning or end, and (3) video ads must include both a 4-second visual statement and a 3-second spoken statement, all in the same language as the communication.

Advertisers who use synthetic media in election campaign communications must include a disclosure stating 'This (image/video/audio) has been manipulated' — for visual media, in a font not smaller than the largest other text and displayed for the video's full duration; for audio-only media, read clearly at the beginning, end, and at intervals of no more than two minutes.

Advertisers who use synthetic media in state or local election campaign communications must include a disclosure stating 'This (image/video/audio) has been manipulated' — for visual media, in a font not smaller than the largest other text and displayed for the video's full duration; for audio-only media, read clearly at the beginning, end, and at intervals of no more than two minutes.

Candidates, campaign finance entities, registered lobbyists or political committees, and their agents must include a prescribed disclosure statement on any campaign material that uses or contains synthetic media — with format-specific requirements: still images must carry visible text at least as large as the largest other text; audio recordings must include a spoken statement at the beginning, end, and at two-minute intervals; and video recordings must display persistent on-screen text for the full duration.

Candidates, campaign finance entities, registered lobbyists, political committees, and their agents must include a prescribed disclosure statement on any campaign material that uses or contains synthetic media — with medium-specific formatting: visible text on still images (type size at least as large as the largest other text), spoken disclosure at the beginning, end, and every two minutes for audio recordings, and persistent on-screen text for video recordings.

Persons who publish synthetic media as a still image must include the prescribed statement that the image has been altered through computer programs, written in a type size easily readable by the average viewer.

Persons who publish synthetic media as an audio recording must include the prescribed spoken disclosure at the beginning and end, and every two minutes for recordings of two minutes or longer, in a clearly spoken manner at an easily audible pitch and speed.

Persons who publish synthetic media as a video recording must include the prescribed disclosure statement as persistent on-screen text visible for the full duration, in an easily readable type size at a conspicuous location in the frame.

Any person or entity required to provide a campaign disclosure must additionally disclose, using prescribed verbatim language, when a political communication contains synthetic media — images, audio, or video that have been materially manipulated or altered using digital technology or other means. Exemptions apply to bona fide news broadcasts, paid broadcasts made in good faith, news publications that disclaim authenticity, and satire or parody.

Any person who creates or distributes a prerecorded telephone message that advocates for the election or defeat of a candidate or on a ballot question, and that is generated in whole or substantially by AI, must include the verbatim disclaimer: This message was generated in whole or in part by artificial intelligence.

Any person, committee, or other entity that creates, originally publishes, or originally distributes a qualified political advertisement must include a clear and conspicuous disclaimer disclosing AI use, with format-specific requirements: (1) text/graphic ads must identify what was AI-produced in matching-size and matching-language text; (2) audio ads must include a spoken disclaimer of at least 3 seconds at the beginning or end; (3) video-with-audio ads must include both a 4-second visual text message and a 3-second spoken disclaimer, all in the same language as the communication.

Any person responsible for a prerecorded telephone message that advocates for or against a candidate or ballot question must include the verbatim disclaimer This message was generated in whole or in part by artificial intelligence if the message was generated in whole or substantially by AI.

Any person, committee, or entity that creates, originally publishes, or originally distributes a qualified political advertisement must include a clear and conspicuous AI-disclosure disclaimer in the applicable format: (1) for text/graphic communications, a written disclaimer identifying what was AI-generated, at least as large as the majority text and in the same language; (2) for audio communications, a spoken disclaimer at the beginning or end lasting at least three seconds in the same language; (3) for video with audio, both a visual message displayed for at least four seconds and a spoken message lasting at least three seconds, in the same language.

Any person who creates or causes distribution of a prerecorded telephone message that advocates for or against a candidate or ballot question and that is generated in whole or substantially by AI must include the verbatim disclaimer: This message was generated in whole or in part by artificial intelligence.

Persons, committees, or entities that create, originally publish, or originally distribute a qualified political advertisement must include a clear and conspicuous AI-disclosure disclaimer in the applicable format: (1) for text/graphic ads, a written disclaimer identifying what was AI-produced, in letters at least as large as the majority text and in the same language; (2) for audio ads, a spoken disclaimer at the beginning or end lasting at least three seconds in the same language; (3) for video-with-audio ads, both a visual disclaimer displayed for at least four seconds and a spoken disclaimer of at least three seconds, in the same language.

Political parties and political candidates using automatic dialing and recorded message players must satisfy the AI disclaimer requirements of G.S. 163-278.18A as a condition of the robocall exemption.

Any person who creates a political advertisement in whole or in part using artificial intelligence must include a disclaimer stating: This advertisement was created using artificial intelligence.

When a political advertisement using AI is transmitted via social media, the disclaimer must comply with the size requirements of G.S. 163-278.39(b); when transmitted via automated calling, the disclaimer must be spoken clearly for at least two seconds.

Creators must not sponsor, create, or distribute a deceptive and fraudulent deepfake of a candidate or political party within 90 days of an election unless the communication includes a prescribed AI-generated content disclosure — audible statements of Contains content generated by AI for audio and persistent on-screen text identifying the AI-generated content type for visual media, plus a legible This (image, video, or audio) has been manipulated statement.

Any person who distributes or publishes a political communication containing materially deceptive media and who knows or should know it is materially deceptive must include a legible This (image, video, or audio) has been manipulated disclosure for visual media, or a spoken statement at the beginning, end, and every two minutes for audio media. Exemptions apply for satire or parody, bona fide news reporting with disclosure, good-faith platform verification, and interactive computer services under 47 U.S.C. § 230.

Any person distributing political communications or advertising that uses AI to visually or audibly impersonate a human must prominently display the disclaimer THIS CONTENT GENERATED BY ARTIFICIAL INTELLIGENCE. This does not apply to content using AI solely for text generation, grammar correction, spelling checks, stylistic editing, or enhancing existing content without creating a new impersonation of human likeness or voice.

Persons who create, originally publish, or originally distribute a qualified political advertisement must include a clear and conspicuous disclosure stating the ad was generated in whole or substantially by AI, with format-specific requirements: minimum text size and language matching for graphic ads; a three-second spoken disclosure for audio ads; and both a four-second on-screen text disclosure and three-second spoken disclosure for video ads. Each distributed or aired ad that violates the disclosure requirement constitutes a separate violation. Exceptions apply for news broadcasts with acknowledgment, paid broadcast stations, distribution platforms with written disclosure policies, satire/parody, and PSC/FCC-regulated businesses.

Any person or committee responsible for a political communication covered by Election Law § 14-106 that was produced using or includes synthetic media must include a disclosure stating This political communication was created with the assistance of artificial intelligence — printed or typed in legible form for printed and digital communications, and clearly and prominently displayed and/or spoken for non-printed communications.

Any person making a political communication that uses AI-generated images or video footage must include a disclosure stating 'This communication was generated using artificial intelligence' — in legible printed form for print/digital media, prominently displayed and/or verbally communicated for non-print/non-digital media, and clearly verbally communicated for audio-only formats such as radio or robocalls.

Any person or committee that produces or distributes a political communication containing synthetic media must include the disclaimer: This political communication was created with the assistance of artificial intelligence — printed in legible form for printed or digital communications, and clearly spoken for non-visual communications such as radio or automated telephone calls.

Any person who knowingly creates and disseminates deepfake media to influence an election must include a conspicuous manipulation disclosure: a written statement for images and video (in a font no smaller than the largest accompanying text, displayed for the full duration for video) and a spoken statement for audio (at the beginning and end, and every two minutes for recordings over four minutes).

Candidates, candidate committees, political action committees, and political party committees that create or distribute a political advertisement, electioneering communication, or other election-related media containing AI-generated content that appears to depict a real person performing an action that did not occur in reality must prominently include the disclosure: Created in whole or in part with the use of generative artificial intelligence. For visual media, the disclosure text must appear in a size easily readable by the average viewer; for video, the disclosure must appear for the full duration of the AI-generated content. For audio-only media, the disclosure must be read in a clearly spoken manner and at a pitch easily heard by the average listener at both the beginning and the end of the audio. Exceptions apply to: (1) bona fide news broadcasts that acknowledge authenticity questions, (2) paid broadcasts where the station made a good-faith verification effort, (3) news publications that clearly state the media does not accurately represent the candidate's speech or conduct, and (4) satire or parody.

Persons distributing synthetic media that is a deepfake of a candidate or political party within 90 days of an election must include a disclosure stating that the content has been manipulated or generated by artificial intelligence, using prescribed formatting: for visual media, text at least as large as the largest other text and displayed for the full duration of video; for audio-only media, read clearly at the beginning, end, and at intervals of no more than two minutes.

Creators who distribute synthetic media depicting a candidate within the 90-day election window must include a disclosure stating the media was manipulated or AI-generated — for visual media in a font at least as large as the largest other text and displayed for the duration of any video, and for audio-only media spoken clearly at the beginning, end, and at two-minute intervals.

Any person financing a political advertisement that contains AI-generated or AI-materially-altered images, audio, or video must include a clear and conspicuous disclaimer stating that the communication contains such content. For images the disclaimer must be permanently affixed in text at least as large as the majority of the communication's text; for audio the disclaimer must be spoken for at least 4 seconds naming the disbursing person; for video with audio the disclaimer must appear in both written and audible form throughout. A safe harbor excludes minor cosmetic alterations that do not create a fundamentally different understanding.

Persons responsible for political communications covered under FECA Section 318(a) must include a clear and conspicuous statement disclosing that the communication contains AI-generated image or video footage, whenever such content was generated in whole or in part using generative AI. Text/graphic disclaimers must use lettering at least as large as the majority text; audio disclaimers must be spoken audibly for at least 3 seconds; video disclaimers must include both a written statement for at least 4 seconds and an audible statement for at least 3 seconds.

A person who finances a generative A.I. political advertisement must include a prominent disclaimer on the advertisement stating: Created in whole or in part with the use of generative artificial intelligence (A.I.).

Any committee or person sponsoring an electioneering communication containing synthetic media must include a conspicuous disclosure statement reading: 'This message contains synthetic media that has been altered from its original source or artificially generated and may present conduct or speech that did not occur.' Visual media must display the statement in proportionate font (minimum seven points for electronic print); video and audio formats must place the statement at the beginning, and also at the end if the communication exceeds five minutes. Satire and parody are exempt.

Any committee or person sponsoring an electioneering communication containing synthetic media must include a conspicuous disclosure statement using the prescribed language: This message contains synthetic media that has been altered from its original source or artificially generated and may present conduct or speech that did not occur. For visual media, the statement must be in a font proportionate to the advertisement size (minimum 7-point for electronic print media). For video or audio, the disclosure must appear at the beginning, and also at the end if the communication exceeds five minutes.

Sponsors of electioneering communications containing synthetic media may avoid liability by including a disclosure stating This (image/video/audio) has been manipulated — for visual media, in a font size at least as large as the largest other text and displayed for the video's full duration; for audio-only media, read clearly at the beginning and end, and at intervals of no more than two minutes for audio longer than two minutes.

Broadcasting stations and other media must not remove synthetic media disclosures from electioneering communications they broadcast, and must not alter electioneering communication content with actual malice such that it becomes synthetic media.

Persons distributing electioneering communications containing synthetic media that alters a candidate's appearance, action, or speech must include a prominent disclosure stating This (image/video/audio) has been manipulated, formatted to be easily readable or audible by the average viewer or listener, displayed for the full duration of any video and repeated at two-minute intervals for audio exceeding two minutes. The obligation applies only where the distributor acts with actual malice. Exemptions apply for satire or parody, bona fide news broadcasts with authenticity acknowledgments, news publications with accuracy disclaimers, and paid broadcast placements.

Political committees, parties, and similar entities that pay for audio communications containing synthetic media must include the words Contains content generated by AI at both the beginning and the end of the communication.

Political committees, parties, and similar entities that pay for video communications containing synthetic media must display readable, legible, and readily accessible written disclosure throughout the duration of each portion containing synthetic media, using the prescribed label text appropriate to the type of synthetic media present (video-only, audio-only, or both).

No person may distribute synthetic media that the person knows or should know is a deceptive and fraudulent deepfake of a candidate or political party on a state or local ballot within 90 days of an election, unless the media includes the required disclosure statement.

No person may distribute, disseminate, publish, broadcast, transmit, or display a communication concerning a candidate for elective office that includes a deepfake to the candidate's electorate within 60 days of a primary election or 90 days of a general election, if the person knows or recklessly disregards that the candidate did not say or do what is depicted. Exemptions apply for bona fide news coverage, satire and parody, technology providers, interactive computer services, and paid broadcast stations.

No person may distribute, or agree to distribute, a communication containing deceptive synthetic media depicting an individual during the 90-day period preceding an election or primary when the person knows or should know the content is deceptive synthetic media, distribution is without the depicted individual's consent, and distribution is intended to injure a candidate or influence the election result — unless proper disclaimers are included per the safe harbor in subsection (c).

No person may distribute a deceptive and fraudulent deepfake of a candidate or political party within 90 days of an election at which the candidate appears on the ballot, unless the media includes a disclosure that the content was manipulated or generated by AI. For visual media, the disclosure must read this image or video was manipulated or generated by artificial intelligence in easily readable text for the duration of any video. For audio-only media, the disclosure must state this audio was manipulated or generated by artificial intelligence at the beginning and end — and at two-minute intervals if longer than two minutes — in a clearly spoken manner.

Sponsors of electioneering communications must not publish synthetic media that alters a candidate's appearance, action, or speech without including the prescribed manipulation disclosure. A candidate depicted in such synthetic media may obtain injunctive relief and recover general damages, special damages, attorney's fees, and costs.

Persons must not use synthetic media in any election campaign communication through any medium to alter the appearance, action, or speech of a candidate. An affirmative defense applies if the communication includes the prescribed manipulation disclosure.

Persons must not use synthetic media in any election campaign communication through any medium to alter the appearance, action, or speech of a candidate for state or local office. An affirmative defense applies if the communication includes the prescribed manipulation disclosure.

No person may use synthetic media in any election campaign communication through any medium to alter the appearance, action, or speech of a candidate. An affirmative defense applies if the communication includes a conspicuous disclosure stating 'This (image/video/audio) has been manipulated,' formatted per statutory requirements for visual and audio media.

No person may use synthetic media in any election campaign communication through any medium to alter the appearance, action, or speech of a candidate for state or local office. An affirmative defense applies if the communication includes a conspicuous disclosure stating 'This (image/video/audio) has been manipulated,' formatted per statutory requirements for visual and audio media.

No person may disseminate an AI-generated or digitally manipulated deepfake to influence a voter's decision whether or not to cast a vote or whether to vote for or against a candidate or ballot issue, where the deepfake materially misrepresents a fact, the person knows or recklessly disregards that it is a deepfake, and the person intends to deceive and induce voting behavior. Exceptions apply for news media dissemination and satire or parody.

A person must not knowingly disseminate — or contract to disseminate — a deep fake within 90 days before an election, without the depicted individual's consent, with intent to injure a candidate or influence the election result.

Creators must not distribute synthetic media that they know or should know is a deceptive and fraudulent deepfake within 18 weeks of an election, unless the media carries the prescribed AI-generation disclosure label.

No candidate, campaign committee, political party, PAC, referendum committee, individual, or other sponsor may use artificial intelligence in any form of political advertisement, including images, videos, voices, or writings.

No unaffiliated person may create an AI-generated video, writing, voice, or image of a candidate or a candidate's campaign with intent to confuse or interfere with the campaign, subject to exceptions for content that clearly identifies itself as AI-generated and for parody that does not constitute harassment.

Persons must not distribute, or agree to distribute, AI-generated materially deceptive media that falsely depicts a person's speech or conduct within 90 days of an election when the person knows the media is false, intends to harm a candidate's reputation or electoral prospects, and intends to deceive voters — unless the media includes a format-specific disclaimer stating it was manipulated by technical means and depicts speech or conduct that did not occur, and edited media includes a citation to the original source.

No person may include in any political communication covered by Election Law § 14-106 a realistic photo, video, or audio depiction of a candidate — or a person interacting with a candidate — that was created in whole or in part using generative artificial intelligence.

No person may knowingly create and disseminate deepfake media for the purpose of influencing the results of an election during the 90-day period before the election through election day — disclosure does not cure this prohibition.

No person, corporation, committee, or other entity may create or originally distribute synthetic media that it has actual knowledge is a deepfake of a candidate or political party within 90 days of an election at which the candidate appears on the state or local ballot, unless the media includes the prescribed AI-generation disclosure label.

Creators must not distribute synthetic media depicting a candidate within 90 days of an election when the creator knows or should know the media is deceptive, unless the media includes a qualifying AI-manipulation disclosure.

No person, political committee, or other entity may distribute a paid-for political communication containing AI-generated audio that impersonates a federal candidate's voice if the distribution is made with actual malice, is intended to injure the candidate's reputation or deceive voters into voting against the candidate, and occurs within 90 days of a general/special/runoff election or 60 days of a primary/preference election/convention/caucus. A clear, conspicuous, and overt disclaimer that the audio was not spoken by the candidate is a safe harbor.

No person, political committee, or other entity may knowingly distribute materially deceptive AI-generated audio or visual media depicting a candidate for federal office, or in carrying out a federal election activity, with the intent to influence an election or solicit funds. Exceptions apply for media entities that clearly disclose authenticity concerns and for satire or parody.

No person may fraudulently misrepresent themselves as speaking, writing, or otherwise acting for or on behalf of a candidate or political party through the use of content generated in whole or in part with artificial intelligence (generative AI).

No person may use content generated in whole or in part with artificial intelligence (generative AI) to fraudulently solicit contributions by misrepresenting the solicitation as being on behalf of a candidate, political party, or political committee.

Any person — not only candidates and their agents — is prohibited from fraudulently misrepresenting themselves as acting on behalf of any candidate, political party, or real or fictitious political committee or organization, including through AI-generated content.

Any person is prohibited from fraudulently soliciting contributions by impersonating any candidate, political party, or real or fictitious political committee or organization, including through AI-generated content.

No person, political committee, or other entity may knowingly distribute materially deceptive AI-generated audio or visual media of a candidate for Federal office, or in carrying out a Federal election activity, with the intent to influence an election or solicit funds. Exemptions apply for news organizations that disclose authenticity concerns and for satire or parody.

Sponsors of electioneering communications must not use synthetic media depicting a candidate's appearance, speech, or conduct with actual malice. A depicted candidate may obtain injunctive relief and general or special damages.

No person may distribute, or agree to distribute, synthetic media that falsely represents a depicted individual within 90 days before an election, without the depicted individual's consent, if the person knows or has reasonable cause to know the media is false and acts with intent to injure a candidate or influence election results. The prohibition does not apply if the synthetic media includes a qualifying disclaimer under subsection (c).

Persons must not disseminate misleading synthetic media with malicious intent, to influence the outcome of any Wyoming election, or to facilitate any other criminal act; violations constitute a misdemeanor punishable by up to six months' imprisonment, a $750 fine, or both.

Operators must not knowingly and intentionally cause or program a conversational AI service to make any representation or statement that explicitly indicates the service is designed to provide professional mental or behavioral health care.

Any person or entity that develops or deploys AI in Missouri must not advertise or represent to the public that the AI is or is able to act as a mental health professional, or is capable of providing therapy services, psychotherapy services, or a mental health diagnosis.

AI providers must not represent, or cause their AI systems to represent, that the system is capable of providing professional mental or behavioral health care, that a user may obtain such care through conversational AI features, or that the system is a provider of such care.

Operators must not knowingly and intentionally cause or program a conversational AI service to make any representation or statement that explicitly indicates the service is designed to provide professional mental or behavioral health care.

A person must not use a price or wage setting algorithm (PWSA) or its output to set individualized prices for consumers based on surveillance data — data derived from observation, inference, or surveillance of personal characteristics, online behaviors, or biometrics. Nine safe harbors apply: (1) cost-justified price differentials based on consumer selections, delivery distance, or delivery time; (2) temporal price fluctuations based on supply and demand; (3) publicly disclosed group discounts offered on equal terms to all consumers meeting published eligibility criteria or to broadly defined publicly recognized groups; (4) loyalty, membership, or rewards program pricing on publicly disclosed equal terms; (5) insurer pricing using only risk-relevant data in compliance with C.R.S. § 10-3-1104.9; (6) good-faith credits, refunds, rebates, or discounts below a previously offered price in response to complaints, service disruptions, billing disputes, retention requests, or other customer service interactions; (7) established need-based discount programs (hospital discounted care, sliding scale fees, charity care) with publicly disclosed objective eligibility criteria related to income or financial need; (8) subscription or continuous-agreement recurring pricing not informed by a PWSA; and (9) credit refusals or financial transaction refusals based on FCRA-covered consumer report data or data required as part of the financial transaction application.

A person must not use a price or wage setting algorithm (PWSA) or its output to set individualized wages for workers based on surveillance data. A safe harbor applies where the person can demonstrate both that (1) wages are based solely on data specific to the individual worker that is directly related to the tasks the worker was hired to perform (including performance data and generation of new business), and (2) the person discloses in plain language before hiring — and to all workers whose wages are set in whole or in part through a PWSA — what data is considered and how the PWSA considers the data. Disclosures that duplicate those required under equal pay, wage and hour, or other existing laws may be satisfied by complying with applicable law.

A person may not engage in surveillance pricing — that is, may not charge, offer, or accept payment of an increased price for goods or services where the price is personalized to a specific consumer or group of consumers based in whole or in part on covered information collected through targeted pricing technologies (sensors, cameras, device tracking, biometric monitoring, cookies, or similar observation), or through real-time AI-driven dynamic pricing that adjusts prices based on customer behavior or other personal attributes. The prohibition does not apply to refusals to extend specific credit terms based on a consumer report obtained in compliance with the federal Fair Credit Reporting Act.

Businesses selling essential goods and services in North Carolina must not engage in surveillance pricing — i.e., must not vary prices within the same business day based on demand or other factors, including through algorithmic pricing that uses consumers' personal data to set individualized prices. Exceptions apply for promotional or loyalty pricing, cost-based price differences, defined group discounts, pricing error corrections, and post-outage price resets.

Landlords must not set rental prices based on recommendations from an AI system that generates rent price recommendations using nonpublic information provided by multiple landlords about their rental prices.

Retail sellers must not offer or sell consumer goods at a price established in whole or in part through dynamic price gouging — i.e., using any computational system or model, including AI, that processes consumer data (data broker purchases, surveillance-obtained data, web-scraped data, data acquired without consent, or protected-class data) to set a customized price for a specific consumer.

Grocery retailers must not engage in personalized algorithmic pricing that results in an essential consumer good being advertised, offered, or sold at a price that varies based on the consumer's sex, race, color, creed, disability, sexual orientation, national origin, or ancestry.

Grocery retailers must not use personalized algorithmic pricing that results in essential consumer good prices varying based on the consumer's sex, race, color, creed, disability, sexual orientation, national origin, or ancestry.

A person must not use covered information (personally identifiable information or aggregate consumer information) gathered through electronic surveillance technology to set the price of a commercial good or service for a consumer at a level that differs from the standard price. Three exceptions apply: (1) the price difference is based solely on the cost of providing the good or service to that consumer; (2) the price difference is a discount offered to all consumers on equal terms consistent with antidiscrimination laws — if personally identifiable information is gathered in connection with the discount, the person must provide clear and conspicuous written notice of intended uses and obtain written affirmative consent before using the information; or (3) the person is an insurer complying with Insurance Code § 791.02 and the pricing is in connection with that activity.

Persons must not engage in surveillance pricing — offering, charging, or accepting payment of an increased price for goods or services where the price is personalized based, in whole or in part, on covered information collected through targeted pricing technologies (including sensors, cameras, biometric monitoring, cookies, device tracking, hardware state, payment-method observation) or through real-time dynamic pricing keyed to customer behavior. Refusals to extend credit or enter transactions based on a consumer report in accordance with the federal Fair Credit Reporting Act are exempt.

No person may use surveillance-based price setting or use surveillance data as part of an automated decision system to set individualized consumer prices for goods or services. Exceptions apply for cost-based price differences, publicly disclosed group discounts, and affirmatively enrolled loyalty or rewards programs.

Landlords must not use any algorithm, commercial computer program, digital technology, or artificial intelligence technology to set or determine the rent for residential property.

Landlords must not use any algorithm, commercial computer program, digital technology, or artificial intelligence to set or determine rent for residential property.

Grocery retailers must not purchase or otherwise acquire from a data broker personal information about a consumer for purposes related to the pricing of any essential consumer good advertised, offered, or sold to that consumer.

Grocery retailers must not electronically display in a grocery store, within the same field of vision as an essential consumer good, a price established through personalized algorithmic pricing.

Grocery retailers must not purchase or otherwise acquire personal information about a consumer from a data broker for purposes related to the pricing of essential consumer goods.

Persons with $5,000,000 or more in annual revenue that use a pricing algorithm to set or recommend prices or commercial terms must clearly disclose to customers (before purchase) and to current or prospective employees or independent contractors that the price or term is algorithmically set or recommended, including whether the algorithm sets different prices for similarly situated customers or workers and the identity of any third-party algorithm developer or distributor.

Persons who knowingly advertise, promote, label, or publish any statement, display, image, offer, or announcement of surveillance pricing must include with that material a clear and conspicuous disclosure stating verbatim: THIS PRICE WAS SET BY A SURVEILLANCE METHOD USING YOUR PERSONAL CONSUMER INFORMATION.

Persons with $5,000,000 or more in annual revenue must clearly disclose to customers before purchase that the price or commercial term is set or recommended by a pricing algorithm, and must disclose to current or prospective employees or independent contractors that their compensation or commercial terms are algorithmically set.

Persons using pricing algorithms must additionally disclose, where applicable, that the algorithm sets or recommends different prices or commercial terms for different customers seeking identical or nearly identical products or services, or for employees or independent contractors providing substantially similar services.

Persons using a pricing algorithm developed or distributed by a third party must disclose that fact and identify the third-party developer or distributor as part of the required algorithmic pricing disclosure.

Grocery retailers must accompany any algorithmically personalized price for an essential consumer good with the verbatim disclosure This price was set by an algorithm using personal data in the same size, format, color, and field of vision as the price. Grocery retailers may not electronically display algorithmically personalized prices in a grocery store within the same field of vision as the product.

Grocery retailers must accompany any algorithmically personalized price for an essential consumer good with the verbatim disclosure This price was set by an algorithm using personal data in the same size, format, color, and field of vision as the price.