{"data":{"records":[{"id":"eu-ai-act-art50-1-chatbot","dataset":"ai-disclosure","jurisdiction":"EU","instrument":"EU AI Act, Regulation (EU) 2024/1689, Art. 50(1)","status":"enacted-not-yet-in-force","summary":"Providers of AI systems intended to interact directly with natural persons must design them so the persons concerned are informed they are interacting with an AI system, unless this is obvious from the point of view of a reasonably well-informed, observant and circumspect natural person.","source_url":"https://artificialintelligenceact.eu/article/50/","source_status":"verified","not_live_law":false,"detail":{"obligation_id":"eu-ai-act-art50-1-chatbot","instrument":"EU AI Act, Regulation (EU) 2024/1689, Art. 50(1)","in_force_date":"2026-08-02","status":"enacted-not-yet-in-force","applies_to":"provider","content_type":"chatbot","trigger":"Providers of AI systems intended to interact directly with natural persons must design them so the persons concerned are informed they are interacting with an AI system, unless this is obvious from the point of view of a reasonably well-informed, observant and circumspect natural person.","required_action":"Disclosure at or before first interaction that the person is dealing with an AI system.","exemptions":"Where obvious to a reasonable person; AI systems authorised by law to detect, prevent, investigate or prosecute criminal offences.","penalty_max":"15000000 or 3% of total worldwide annual turnover, whichever higher (Art. 99(4))","penalty_currency":"EUR","enforced_yet":"false","source_type":"legislative-text","notes":"Not yet in force as of 2026-07-28; bites 2026-08-02. Not deferred by the Digital Omnibus. Per Bird & Bird, the European Commission adopted the final 51-page Article 50 Guidelines on 20 July 2026."}},{"id":"eu-ai-act-art50-2-marking","dataset":"ai-disclosure","jurisdiction":"EU","instrument":"EU AI Act, Regulation (EU) 2024/1689, Art. 50(2)","status":"enacted-not-yet-in-force","summary":"Providers of AI systems, including general-purpose AI, generating synthetic audio, image, video or text content shall ensure the outputs are marked in a machine-readable format and detectable as artificially generated or manipulated.","source_url":"https://artificialintelligenceact.eu/article/50/","source_status":"verified","not_live_law":false,"detail":{"obligation_id":"eu-ai-act-art50-2-marking","instrument":"EU AI Act, Regulation (EU) 2024/1689, Art. 50(2)","in_force_date":"2026-08-02","status":"enacted-not-yet-in-force","applies_to":"provider","content_type":"synthetic-media","trigger":"Providers of AI systems, including general-purpose AI, generating synthetic audio, image, video or text content shall ensure the outputs are marked in a machine-readable format and detectable as artificially generated or manipulated.","required_action":"Machine-readable marking/watermark or provenance metadata, effective, interoperable, robust and reliable as far as technically feasible.","exemptions":"Assistive functions for standard editing that do not substantially alter input data or its semantics; systems authorised by law for criminal-offence detection.","penalty_max":"15000000 or 3% of total worldwide annual turnover, whichever higher","penalty_currency":"EUR","enforced_yet":"false","source_type":"legislative-text","notes":"Systems already placed on the EU market before 2026-08-02 have until 2026-12-02 to comply with the marking obligation per the Digital Omnibus provisional agreement of 7 May 2026. A voluntary Code of Practice on Transparency of AI-Generated Content (first draft 17 Dec 2025) has been assessed as adequate by the Commission."}},{"id":"eu-ai-act-art50-4-deepfake","dataset":"ai-disclosure","jurisdiction":"EU","instrument":"EU AI Act, Regulation (EU) 2024/1689, Art. 50(4) first subparagraph","status":"enacted-not-yet-in-force","summary":"Deployers of an AI system that generates or manipulates image, audio or video content constituting a deep fake shall disclose that the content has been artificially generated or manipulated.","source_url":"https://artificialintelligenceact.eu/article/50/","source_status":"verified","not_live_law":false,"detail":{"obligation_id":"eu-ai-act-art50-4-deepfake","instrument":"EU AI Act, Regulation (EU) 2024/1689, Art. 50(4) first subparagraph","in_force_date":"2026-08-02","status":"enacted-not-yet-in-force","applies_to":"deployer","content_type":"deepfake","trigger":"Deployers of an AI system that generates or manipulates image, audio or video content constituting a deep fake shall disclose that the content has been artificially generated or manipulated.","required_action":"Clear, perceptible disclosure (visible/audible label) at the latest at the time of first exposure; deployers cannot rely solely on the provider's machine-readable mark.","exemptions":"Use authorised by law to detect/prevent/investigate/prosecute crime; for evidently artistic, creative, satirical or fictional works the duty is limited to disclosure in a manner that does not hamper display or enjoyment of the work.","penalty_max":"15000000 or 3% of total worldwide annual turnover, whichever higher","penalty_currency":"EUR","enforced_yet":"false","source_type":"legislative-text","notes":"'Deep fake' defined in Art. 3(60). Deployer duty distinct from the provider marking duty in 50(2)."}},{"id":"eu-ai-act-art50-4-public-text","dataset":"ai-disclosure","jurisdiction":"EU","instrument":"EU AI Act, Regulation (EU) 2024/1689, Art. 50(4) second subparagraph","status":"enacted-not-yet-in-force","summary":"Deployers of an AI system that generates or manipulates text which is published with the purpose of informing the public on matters of public interest shall disclose that the text has been artificially generated or manipulated.","source_url":"https://artificialintelligenceact.eu/article/50/","source_status":"verified","not_live_law":false,"detail":{"obligation_id":"eu-ai-act-art50-4-public-text","instrument":"EU AI Act, Regulation (EU) 2024/1689, Art. 50(4) second subparagraph","in_force_date":"2026-08-02","status":"enacted-not-yet-in-force","applies_to":"deployer","content_type":"text","trigger":"Deployers of an AI system that generates or manipulates text which is published with the purpose of informing the public on matters of public interest shall disclose that the text has been artificially generated or manipulated.","required_action":"Disclosure that the text is AI-generated/manipulated.","exemptions":"Where the AI-generated content has undergone a process of human review or editorial control and a natural or legal person holds editorial responsibility; law-enforcement authorised uses.","penalty_max":"15000000 or 3% of total worldwide annual turnover, whichever higher","penalty_currency":"EUR","enforced_yet":"false","source_type":"legislative-text","notes":"The editorial-control carve-out is the key escape for newsrooms."}},{"id":"us-ca-ab2013-training-data","dataset":"ai-disclosure","jurisdiction":"US-state (California)","instrument":"California AB 2013 (Generative AI Training Data Transparency Act), Cal. Civ. Code § 3110 et seq.","status":"in-force","summary":"On or before 1 January 2026, before making a generative AI system/service (or a substantial modification) released on or after 1 January 2022 publicly available to Californians, a developer must post documentation on its website regarding the data used to train the system.","source_url":"https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202320240AB2013","source_status":"verified","not_live_law":false,"detail":{"obligation_id":"us-ca-ab2013-training-data","instrument":"California AB 2013 (Generative AI Training Data Transparency Act), Cal. Civ. Code § 3110 et seq.","in_force_date":"2026-01-01","status":"in-force","applies_to":"provider","content_type":"any","trigger":"On or before 1 January 2026, before making a generative AI system/service (or a substantial modification) released on or after 1 January 2022 publicly available to Californians, a developer must post documentation on its website regarding the data used to train the system.","required_action":"Public website disclosure of a high-level summary of training datasets (sources, whether copyrighted/licensed, whether personal information or synthetic data included, time period, etc.).","exemptions":"AI systems used solely for security and integrity, aircraft/national airspace operation, or national security/defence purposes.","enforced_yet":"false","source_type":"legislative-text","notes":"SCOPE CAVEAT: this is TRAINING-DATA transparency, not a duty to label individual AI outputs as AI-generated. No express penalty, trade-secret exemption, or prescribed format in the statute. Enforced by the California AG under general authority; early attention is on developers that failed to publish at all."}},{"id":"us-ca-sb942-provider-disclosure","dataset":"ai-disclosure","jurisdiction":"US-state (California)","instrument":"California AI Transparency Act, Cal. Bus. & Prof. Code §§ 22757.3–22757.30 (SB 942, as amended by AB 853)","status":"enacted-not-yet-in-force","summary":"A 'covered provider' (creates/codes/produces a generative AI system with more than 1,000,000 monthly visitors or users that is publicly accessible in California) must include disclosures in image, video or audio content its system creates or alters.","source_url":"https://leginfo.legislature.ca.gov/faces/billNavClient.xhtml?bill_id=202320240SB942","source_status":"verified","not_live_law":false,"detail":{"obligation_id":"us-ca-sb942-provider-disclosure","instrument":"California AI Transparency Act, Cal. Bus. & Prof. Code §§ 22757.3–22757.30 (SB 942, as amended by AB 853)","in_force_date":"2026-08-02","status":"enacted-not-yet-in-force","applies_to":"provider","content_type":"synthetic-media","trigger":"A 'covered provider' (creates/codes/produces a generative AI system with more than 1,000,000 monthly visitors or users that is publicly accessible in California) must include disclosures in image, video or audio content its system creates or alters.","required_action":"Provide a free public AI-detection tool; embed a latent (machine-readable) disclosure in AI-generated content (provider name, system name/version, time of creation, unique identifier); offer an option for a manifest (visible) disclosure.","exemptions":"Providers below the 1,000,000-user threshold; product features/content not substantially generated by the system.","penalty_max":"5000 per violation, each day of violation a discrete violation (Cal. Bus. & Prof. Code § 22757.30)","penalty_currency":"USD","enforced_yet":"false","source_type":"legislative-text","notes":"Statute: 'A covered provider that violates this chapter shall be liable for a civil penalty in the amount of five thousand dollars ($5,000) per violation... Each day that a covered provider is in violation of this chapter shall be deemed a discrete violation.' AB 853 (signed 13 Oct 2025, Ch. 674 Stats. 2025) delayed the operative date from 2026-01-01 to 2026-08-02 to align with the EU AI Act. Enforced by AG / city/county attorneys."}},{"id":"us-ca-ab853-hosting-platform","dataset":"ai-disclosure","jurisdiction":"US-state (California)","instrument":"California AB 853 (AI Transparency Act amendments), Ch. 674 Stats. 2025 (amending Cal. Bus. & Prof. Code § 22757 et seq.)","status":"enacted-not-yet-in-force","summary":"From 1 January 2027, generative AI system hosting platforms (websites/apps that make available for download the source code or model weights of a generative AI system) become subject to the Act, and covered providers must not make available a generative AI system lacking the required transparency disclosures.","source_url":"https://hintzelaw.com/blog/2025/10/19/california-amends-artificial-intelligence-transparency-act-and-passes-ai-defenses-act","source_status":"verified","not_live_law":false,"detail":{"obligation_id":"us-ca-ab853-hosting-platform","instrument":"California AB 853 (AI Transparency Act amendments), Ch. 674 Stats. 2025 (amending Cal. Bus. & Prof. Code § 22757 et seq.)","in_force_date":"2027-01-01","status":"enacted-not-yet-in-force","applies_to":"platform","content_type":"synthetic-media","trigger":"From 1 January 2027, generative AI system hosting platforms (websites/apps that make available for download the source code or model weights of a generative AI system) become subject to the Act, and covered providers must not make available a generative AI system lacking the required transparency disclosures.","required_action":"Hosting-platform transparency/provenance obligations; provider prohibition on distributing non-compliant systems.","exemptions":"Defined-term thresholds for 'large online platform' (>2,000,000 unique monthly California users) and 'capture device manufacturer'.","penalty_currency":"USD","enforced_yet":"false","source_type":"regulator-guidance","notes":"Exact newly-added code sections for the large-online-platform and capture-device duties, and their later phased dates (reported to run through 2028), could not be confirmed from primary text; see gaps section. Best available primary-adjacent summary cited."}},{"id":"us-tx-traiga-govt-disclosure","dataset":"ai-disclosure","jurisdiction":"US-state (Texas)","instrument":"Texas Responsible AI Governance Act (HB 149), Tex. Bus. & Com. Code § 552.051","status":"in-force","summary":"A governmental agency that makes available an AI system intended to interact with consumers must disclose to the consumer, before or at the time of the interaction, that the consumer is interacting with an AI system.","source_url":"https://www.nortonrosefulbright.com/en/knowledge/publications/c6c60e0c/the-texas-responsible-ai-governance-act","source_status":"verified","not_live_law":false,"detail":{"obligation_id":"us-tx-traiga-govt-disclosure","instrument":"Texas Responsible AI Governance Act (HB 149), Tex. Bus. & Com. Code § 552.051","in_force_date":"2026-01-01","status":"in-force","applies_to":"deployer","content_type":"chatbot","trigger":"A governmental agency that makes available an AI system intended to interact with consumers must disclose to the consumer, before or at the time of the interaction, that the consumer is interacting with an AI system.","required_action":"Clear and conspicuous disclosure before or at the point of interaction (plain language, not solely buried in terms of service).","exemptions":"Applies to governmental entities; private employers are not required to disclose AI use to applicants/employees.","penalty_max":"200000 per uncurable violation (civil penalties $10,000–$12,000 curable; $80,000–$200,000 uncurable; $2,000–$40,000/day continuing)","penalty_currency":"USD","enforced_yet":"false","source_type":"regulator-guidance","notes":"Signed 22 Jun 2025; effective 1 Jan 2026. Enforced exclusively by the Texas AG (60-day cure period; no private right of action). TRAIGA also requires health-care providers to disclose to patients when AI is used in their treatment (exact section not independently confirmed). Best available primary-adjacent summary cited pending confirmation of the codified section."}},{"id":"us-ut-aipa-genai-disclosure","dataset":"ai-disclosure","jurisdiction":"US-state (Utah)","instrument":"Utah AI Policy Act, Utah Code Ann. Title 13, Ch. 72 (SB 149, as amended by SB 226 & SB 332)","status":"in-force","summary":"A person who uses generative AI to interact with a consumer must clearly and conspicuously disclose that the consumer is interacting with generative AI, and not a human, if the consumer makes 'a clear and unambiguous request' to know; proactive disclosure is required in 'high-risk' interactions (collection of health, financial or biometric data, or personalized legal/medical/financial advice) and by persons in regulated occupations.","source_url":"https://fpf.org/blog/chatbots-in-check-utahs-latest-ai-legislation/","source_status":"verified","not_live_law":false,"detail":{"obligation_id":"us-ut-aipa-genai-disclosure","instrument":"Utah AI Policy Act, Utah Code Ann. Title 13, Ch. 72 (SB 149, as amended by SB 226 & SB 332)","in_force_date":"2025-05-07","status":"in-force","applies_to":"deployer","content_type":"chatbot","trigger":"A person who uses generative AI to interact with a consumer must clearly and conspicuously disclose that the consumer is interacting with generative AI, and not a human, if the consumer makes 'a clear and unambiguous request' to know; proactive disclosure is required in 'high-risk' interactions (collection of health, financial or biometric data, or personalized legal/medical/financial advice) and by persons in regulated occupations.","required_action":"Clear and conspicuous disclosure of the non-human/AI nature of the interaction.","exemptions":"Safe harbour (SB 226): no enforcement if the GenAI itself clearly and conspicuously discloses it is not human at the outset of and throughout the interaction.","penalty_max":"2500 per violation (administrative); up to $5,000 per violation of a court/administrative order","penalty_currency":"USD","enforced_yet":"false","source_type":"regulator-guidance","notes":"Original SB 149 in force 1 May 2024; SB 226/SB 332 amendments effective 7 May 2025 narrowed proactive disclosure to on-request + high-risk and extended the sunset to 1 July 2027. Enforced by the Utah Division of Consumer Protection / AG. Exact disclosure sub-section within Ch. 72 not independently confirmed to sub-section granularity."}},{"id":"cn-deep-synthesis-art17","dataset":"ai-disclosure","jurisdiction":"China","instrument":"Provisions on the Administration of Deep Synthesis of Internet Information Services (2022), Art. 17 (with Art. 16)","status":"in-force","summary":"Deep-synthesis service providers whose services generate or significantly edit content (e.g., intelligent dialogue, synthetic voice, face generation/swapping, immersive scenes) that may cause public confusion or mistaken identification must add conspicuous labels in prominent positions so the public knows the content is synthesized.","source_url":"https://www.loc.gov/item/global-legal-monitor/2023-04-25/china-provisions-on-deep-synthesis-technology-enter-into-effect/","source_status":"blocked","not_live_law":false,"detail":{"obligation_id":"cn-deep-synthesis-art17","instrument":"Provisions on the Administration of Deep Synthesis of Internet Information Services (2022), Art. 17 (with Art. 16)","in_force_date":"2023-01-10","status":"in-force","applies_to":"provider","content_type":"synthetic-media","trigger":"Deep-synthesis service providers whose services generate or significantly edit content (e.g., intelligent dialogue, synthetic voice, face generation/swapping, immersive scenes) that may cause public confusion or mistaken identification must add conspicuous labels in prominent positions so the public knows the content is synthesized.","required_action":"Conspicuous visible label in a prominent location on the generated/edited content; Art. 16 requires an embedded, non-removable technical mark.","exemptions":"Deep-synthesis services outside the listed confusion-risk categories need only provide a labelling function that users can invoke.","enforced_yet":"false","source_type":"legislative-text","notes":"Issued by CAC/MIIT/MPS 25 Nov 2022, effective 10 Jan 2023. Penalties handled under the Cybersecurity Law and other regulations; no fixed figure in the Provisions."}},{"id":"cn-genai-interim-art12","dataset":"ai-disclosure","jurisdiction":"China","instrument":"Interim Measures for the Management of Generative AI Services (2023), Art. 12","status":"in-force","summary":"Providers must label AI-generated pictures, videos and other content in accordance with the Deep Synthesis Provisions.","source_url":"https://www.whitecase.com/insight-our-thinking/ai-watch-global-regulatory-tracker-china","source_status":"blocked","not_live_law":false,"detail":{"obligation_id":"cn-genai-interim-art12","instrument":"Interim Measures for the Management of Generative AI Services (2023), Art. 12","in_force_date":"2023-08-15","status":"in-force","applies_to":"provider","content_type":"synthetic-media","trigger":"Providers must label AI-generated pictures, videos and other content in accordance with the Deep Synthesis Provisions.","required_action":"Labelling of generated content per the Deep Synthesis Provisions.","exemptions":"Measures apply to services offered to the public in mainland China; internal/enterprise R&D not offering services to the public is out of scope.","enforced_yet":"false","source_type":"legislative-text","notes":"Promulgated 10 Jul 2023 by CAC and six other bodies, effective 15 Aug 2023. Cross-refers to the Deep Synthesis Provisions for labelling detail. Primary CAC text at cac.gov.cn."}},{"id":"cn-labelling-measures-2025","dataset":"ai-disclosure","jurisdiction":"China","instrument":"Measures for the Labelling of AI-Generated Synthetic Content (2025), Arts. 4–5 & Art. 14; mandatory national standard GB 45438-2025","status":"in-force","summary":"Internet information service providers that generate/synthesize content (text, images, audio, video, virtual scenes) must add both an explicit label (visible/audible indicator informing users the content is AI-generated) and an implicit label (machine-readable metadata/watermark embedded in the file).","source_url":"https://www.chinalawtranslate.com/en/ai-labeling/","source_status":"verified","not_live_law":false,"detail":{"obligation_id":"cn-labelling-measures-2025","instrument":"Measures for the Labelling of AI-Generated Synthetic Content (2025), Arts. 4–5 & Art. 14; mandatory national standard GB 45438-2025","in_force_date":"2025-09-01","status":"in-force","applies_to":"provider","content_type":"any","trigger":"Internet information service providers that generate/synthesize content (text, images, audio, video, virtual scenes) must add both an explicit label (visible/audible indicator informing users the content is AI-generated) and an implicit label (machine-readable metadata/watermark embedded in the file).","required_action":"Explicit visible/audible label + implicit metadata/digital watermark; downloaded/exported files must retain explicit labels; distribution platforms must verify and preserve labels; users publishing such content must proactively declare it.","exemptions":"A user may request unlabelled output where the user agreement clarifies the user's own labelling obligations and the provider retains logs for at least 6 months.","enforced_yet":"false","source_type":"legislative-text","notes":"Art. 14: 'These measures are to take effect from September 1, 2025.' Issued 14 Mar 2025 by CAC/MIIT/MPS/NRTA together with the mandatory standard GB 45438-2025 (SAMR/SAC), effective the same date. Enforcement via internet-information, telecom, public-security and broadcast authorities under existing laws; CAC 'Qinglang' campaigns are active but no named published decision confirmed under these Measures specifically."}},{"id":"kr-ai-basic-act-art31","dataset":"ai-disclosure","jurisdiction":"South Korea","instrument":"AI Basic Act (Framework Act on the Development of AI and Establishment of a Foundation for Trustworthiness), Act No. 20676, Art. 31 (penalties Art. 43)","status":"in-force","summary":"Operators providing products/services using high-impact AI or generative AI must notify users in advance that the product/service is based on such AI (Art. 31(1)); must indicate that outputs are generated by generative AI (Art. 31(2)); and where synthetic audio/image/video outputs are difficult to distinguish from real content, must notify or indicate to users in a clearly recognizable manner that the content was AI-generated (Art. 31(3)).","source_url":"https://www.loc.gov/item/global-legal-monitor/2026-02-20/south-korea-comprehensive-ai-legal-framework-takes-effect","source_status":"blocked","not_live_law":false,"detail":{"obligation_id":"kr-ai-basic-act-art31","instrument":"AI Basic Act (Framework Act on the Development of AI and Establishment of a Foundation for Trustworthiness), Act No. 20676, Art. 31 (penalties Art. 43)","in_force_date":"2026-01-22","status":"in-force","applies_to":"provider","content_type":"any","trigger":"Operators providing products/services using high-impact AI or generative AI must notify users in advance that the product/service is based on such AI (Art. 31(1)); must indicate that outputs are generated by generative AI (Art. 31(2)); and where synthetic audio/image/video outputs are difficult to distinguish from real content, must notify or indicate to users in a clearly recognizable manner that the content was AI-generated (Art. 31(3)).","required_action":"Advance user notice of AI use; label generative-AI outputs (human-recognizable label or machine-readable/watermark for general outputs; clearly user-recognizable label for realistic synthetic content).","exemptions":"For artistic/creative works the indication may be made in a manner that does not hinder their display, appreciation or other uses.","penalty_max":"30000000 administrative fine","penalty_currency":"KRW","enforced_yet":"false","source_type":"legislative-text","notes":"Promulgated 21 Jan 2025, in force 22 Jan 2026. The 30M KRW fine (Art. 43(1)) attaches directly only to failure of the Art. 31(1) prior-notice duty; breaches of 31(2)/(3) are enforced via corrective orders (Art. 40), non-compliance with which is separately finable. Per the US ITA (trade.gov): 'MSIT is operating a grace period of at least one year in 2026, during which fact-finding investigations and administrative fines will generally be deferred except in exceptional cases involving serious social harm, such as loss of life or human-rights violations.' Labelling method detailed in Enforcement Decree Art. 23."}},{"id":"in-it-rules-sgi-2026","dataset":"ai-disclosure","jurisdiction":"India","instrument":"Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2026, Rule 3(3) & Rule 3(1)(b) proviso (G.S.R. 120(E), 10 Feb 2026)","status":"in-force","summary":"Intermediaries providing computer resources/tools that enable creation or modification of 'synthetically generated information' (SGI) must ensure such content is prominently labelled and embedded with a permanent, unique metadata/identifier; significant social-media intermediaries must ensure SGI carries a prominently visible label or prefixed audio disclosure identifying it as synthetically generated.","source_url":"https://www.meity.gov.in/static/uploads/2025/10/065b6deb585441b5ccdf8be42502a49c.pdf","source_status":"blocked","not_live_law":false,"detail":{"obligation_id":"in-it-rules-sgi-2026","instrument":"Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2026, Rule 3(3) & Rule 3(1)(b) proviso (G.S.R. 120(E), 10 Feb 2026)","in_force_date":"2026-02-20","status":"in-force","applies_to":"platform","content_type":"synthetic-media","trigger":"Intermediaries providing computer resources/tools that enable creation or modification of 'synthetically generated information' (SGI) must ensure such content is prominently labelled and embedded with a permanent, unique metadata/identifier; significant social-media intermediaries must ensure SGI carries a prominently visible label or prefixed audio disclosure identifying it as synthetically generated.","required_action":"Prominent visible/audible label + embedded permanent unique metadata identifier; label/metadata must not be modified, suppressed or removed; user declaration for uploaded SGI.","exemptions":"Good-faith removal/disabling of SGI preserves the Section 79 safe harbour; the final Rules dropped the draft's fixed '10% of surface area / 10% of audio duration' quantitative label-size rule.","enforced_yet":"false","source_type":"legislative-text","notes":"Draft released 22 Oct 2025; Notified Rules published 10 Feb 2026, effective 20 Feb 2026 (FINALISED, not merely draft). No fixed monetary penalty; non-compliance costs the intermediary its Section 79 safe harbour. Applies platform-wide, not per-post (per MeitY FAQ)."}},{"id":"us-ca-ab2839-election-deepfake","dataset":"ai-disclosure","jurisdiction":"US-state (California)","instrument":"California AB 2839 (2024), Cal. Elec. Code § 20012","status":"vacated","summary":"Prohibited knowingly distributing an advertisement or other election communication containing 'materially deceptive' audio/visual media of a candidate within specified pre/post-election windows; required exempt parody/satire to be labelled.","source_url":"https://reason.com/volokh/2025/08/29/california-law-restricting-materially-deceptive-election-related-deepfakes-violates-first-amendment/","source_status":"verified","not_live_law":true,"detail":{"obligation_id":"us-ca-ab2839-election-deepfake","instrument":"California AB 2839 (2024), Cal. Elec. Code § 20012","in_force_date":"2024-09-17","status":"vacated","applies_to":"publisher","content_type":"deepfake","trigger":"Prohibited knowingly distributing an advertisement or other election communication containing 'materially deceptive' audio/visual media of a candidate within specified pre/post-election windows; required exempt parody/satire to be labelled.","required_action":"Prohibition plus disclosure/label requirement for exempt satire/parody.","exemptions":"Candidate deepfakes of themselves; labelled satire/parody.","penalty_max":"injunctive relief and general/special damages (private right of action)","penalty_currency":"USD","enforced_yet":"false","source_type":"official-decision","notes":"STRUCK DOWN. Kohls v. Bonta, No. 2:24-cv-02527-JAM-CKD (E.D. Cal.); preliminary injunction Oct 2024, permanent injunction/summary judgment 29 Aug 2025 on First Amendment grounds. State appealed to the Ninth Circuit. Widely still cited as active law — it is not enforceable."}},{"id":"us-ca-ab2655-platform-deepfake","dataset":"ai-disclosure","jurisdiction":"US-state (California)","instrument":"California AB 2655 (2024), Defending Democracy from Deepfake Deception Act, Cal. Elec. Code § 20510 et seq.","status":"vacated","summary":"Required large online platforms to block or label 'materially deceptive' election-related content during specified windows and to provide a reporting/labelling mechanism.","source_url":"https://www.cahill.com/news/firm-news/2025-08-08-cahill-secures-victory-for-global-social-media-company-overturning-californias-anti-deepfake-law","source_status":"verified","not_live_law":true,"detail":{"obligation_id":"us-ca-ab2655-platform-deepfake","instrument":"California AB 2655 (2024), Defending Democracy from Deepfake Deception Act, Cal. Elec. Code § 20510 et seq.","in_force_date":"2025-01-01","status":"vacated","applies_to":"platform","content_type":"deepfake","trigger":"Required large online platforms to block or label 'materially deceptive' election-related content during specified windows and to provide a reporting/labelling mechanism.","required_action":"Block, or label as inauthentic/fake/false, plus notice-and-report procedures.","exemptions":"Satire/parody; qualifying news outlets.","penalty_max":"injunctive relief","penalty_currency":"USD","enforced_yet":"false","source_type":"official-decision","notes":"STRUCK DOWN. X Corp. v. Bonta, consolidated with Kohls v. Bonta; Judge Mendez held AB 2655 preempted by Section 230 of the Communications Decency Act (Aug 2025). Not enforceable."}},{"id":"es-ai-governance-bill-labelling","dataset":"ai-disclosure","jurisdiction":"Spain","instrument":"Proyecto de Ley Orgánica para el Buen Uso y la Gobernanza de la Inteligencia Artificial (Spanish AI Bill)","status":"proposed","summary":"Content generated or manipulated by AI must be clearly labelled as AI-generated 'in a clear and distinguishable manner no later than the time of the first interaction or exposure' (mirroring EU AI Act Art. 50).","source_url":"https://www.whitecase.com/insight-our-thinking/ai-watch-global-regulatory-tracker-spain","source_status":"blocked","not_live_law":true,"detail":{"obligation_id":"es-ai-governance-bill-labelling","instrument":"Proyecto de Ley Orgánica para el Buen Uso y la Gobernanza de la Inteligencia Artificial (Spanish AI Bill)","status":"proposed","applies_to":"provider","content_type":"synthetic-media","trigger":"Content generated or manipulated by AI must be clearly labelled as AI-generated 'in a clear and distinguishable manner no later than the time of the first interaction or exposure' (mirroring EU AI Act Art. 50).","required_action":"Clear labelling of AI-generated/manipulated content.","exemptions":"Aligned with EU AI Act carve-outs (artistic works, etc.); reduced penalties contemplated for start-ups/SMEs.","penalty_max":"35000000 or 7% of global annual turnover for very serious infringements (mislabelling is a 'serious' infringement: EUR 7.5m–35m or 2–7% turnover)","penalty_currency":"EUR","enforced_yet":"false","source_type":"regulator-guidance","notes":"Council of Ministers approved the preliminary draft 11 Mar 2025; approved as an Organic Bill and sent to Parliament 26 May 2026 (published in the Congreso Official Journal 12 Jun 2026). STILL A BILL; requires an absolute majority of the Congress of Deputies. Supervisor: AESIA. No consolidated primary statute text yet."}},{"id":"br-pl2338-art19","dataset":"ai-disclosure","jurisdiction":"Brazil","instrument":"Projeto de Lei nº 2338/2023 (Marco Legal da IA), Art. 19 (with Art. 20)","status":"proposed","summary":"When an AI system generates synthetic content, the content must carry identifiers allowing verification of its authenticity and provenance.","source_url":"https://www25.senado.leg.br/web/atividade/materias/-/materia/157233","source_status":"verified","not_live_law":true,"detail":{"obligation_id":"br-pl2338-art19","instrument":"Projeto de Lei nº 2338/2023 (Marco Legal da IA), Art. 19 (with Art. 20)","status":"proposed","applies_to":"provider","content_type":"synthetic-media","trigger":"When an AI system generates synthetic content, the content must carry identifiers allowing verification of its authenticity and provenance.","required_action":"Embed authenticity/provenance identifiers (marks) in AI-generated synthetic content; public and private sectors to promote identification/labelling capabilities (Art. 20).","exemptions":"To be defined; the Senate substitutive adds reinforced labelling for deepfakes.","penalty_max":"50000000 per infraction (proposed administrative sanction regime)","penalty_currency":"BRL","enforced_yet":"false","source_type":"legislative-text","notes":"Approved by the Federal Senate 10 Dec 2024; as of July 2026 status in the Câmara dos Deputados special committee is 'Aguardando Parecer do(a) Relator(a)' (rapporteur Dep. Aguinaldo Ribeiro, PP-PB) — NOT enacted. Content-identification obligation confirmed as Art. 19; the separate right to be informed one is interacting with an AI sits in the rights chapter but its exact article number is unconfirmed."}}],"total":18,"coverage":{"corpus_total":18,"matched":18,"jurisdictions_in_corpus":["Brazil","China","EU","India","South Korea","Spain","US-state (California)","US-state (Texas)","US-state (Utah)"],"by_status":{"enacted-not-yet-in-force":6,"in-force":8,"vacated":2,"proposed":2},"source_health":{"verified":13,"blocked":5},"dead_rules_included":4},"as_of":"2026-07-28","interpretation":"Reference data, not legal advice. Every row cites a primary source — legislative text, a regulator document, or a court opinion — and records whether that URL resolved when the corpus was built; \"blocked\" means the host refuses automated checkers, not that the citation is bad. Rows marked not_live_law were vacated, superseded, or never took effect, and are included deliberately: a rule wrongly believed to be in force is the most expensive error in this area. Verify against the cited source before relying on any row."},"_meta":{"request_id":"req_bb1e7b7d8808458a8fbb49fdf27d8e00","product":"legal/ai-disclosure","version":"1.0.0","receipt_id":null,"cost":{"quoted":0,"charged":0,"currency":"USD","display":"0.000000"},"cache":{"hit":true,"age_s":29857},"provenance":{"sourcing_class":"B","upstream":["internal:legal-corpus@2026-07"],"fetched_at":"2026-10-13T11:53:06.297Z","license":"proprietary"},"freshness":{"as_of":"2026-07-28","staleness_s":29857},"confidence":1}}