This dossier contains 25 indexed evidence records connected to United Kingdom. Use the filters to inspect status, instrument type, and topic coverage—including deepfakes, synthetic media, AI-generated content, detection, and KYC—without treating a summary as a legal conclusion.
25indexed records
1source files
12indexed topics
Evidence boundary
A registry record is not a legal opinion.
The cards below reproduce the registry's evidence boundary. A partial record stays partial, and an absent field is shown as not recorded. This informational research is not legal advice or a compliance certification. Detection results, labels, watermarks, provenance records, and biometric or KYC controls do not prove authenticity or compliance.
Source files in this dossier
research/regulations/uk.md
Indexed evidence
Regulation records
Showing 25 of 25 records
Active law, partial
Online Safety Act 2023, 2023 c. 50, especially sections 4, 7–22, 25–37, 55–60, 80–82, 131 and 239–240.
Law — active, with staged commencement and continuing delegated implementation.
CitationNot recorded in registry
Covered actors
Providers of regulated user-to-user services.
Providers of regulated search services, including combined services.
Certain internet services publishing or displaying regulated provider pornography.
Category 1 and Category 2A services for additional duties, including fraudulent advertising.
Users and other persons can be affected by the Act’s content and communications offences, but the principal regulatory duties fall on providers.
Neutral effect or obligation summary
Conduct and keep current illegal-content risk assessments and, where applicable, children’s risk assessments and children’s access assessments. Use proportionate systems and processes to prevent users encountering priority illegal content, mitigate risks of priority offences and other illegal-content harms, minimise the time illegal content remains available, and swiftly take down illegal content when alerted or otherwise aware. Maintain clear, accessible terms of service explaining protection from illegal content, apply those terms consistently, and describe relevant proactive technology, including its type, use and operation. Protect freedom of expression and privacy as required by the Act when implementing safety duties. Providers of services likely to be accessed by children must meet additional duties concerning content harmful to children, governance, reporting, design and safety measures. From 2026-06-29, when a qualifying intimate-image-content report is made, a regulated user-to-user provider must take down the reported content and substantially the same content as soon as reasonably practicable and no later than 48 hours, subject to the statutory exceptions.
DatesNot recorded in registry
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
Synthetic MediaArtificial IntelligenceFraud And ImpersonationPrivacy And Data ProtectionPlatform Safety And ContentChild SafetyProvenance And Labeling
Law — active for the statutory duties; Ofcom’s first fraudulent-advertising codes were still at consultation/draft stage on the verification date.
CitationNot recorded in registry
Covered actors
Providers of Category 1 services and Category 2A search services. The categories are determined by statutory and secondary-legislation thresholds
not every platform, search engine or advertiser is Category 1 or 2A.
Neutral effect or obligation summary
Category 1 services must use proportionate systems and processes to prevent users encountering fraudulent advertisements, minimise the time such content is present, and swiftly take it down when alerted or otherwise aware. Category 2A services must use proportionate systems and processes to prevent fraudulent advertisements being encountered in or via search results, minimise the time they can be encountered, and swiftly ensure they are no longer encountered when alerted or otherwise aware. Relevant terms of service must explain proactive technology used for the duty. “Fraudulent advertisement” is limited by the Act to paid-for advertising that amounts to one of the offences in section 40. Section 40 includes specified Financial Services and Markets Act 2000, Fraud Act 2006 and Financial Services Act 2012 offences, including fraud by false representation and false financial-authorisation claims.
DatesNot recorded in registry
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
Synthetic MediaArtificial IntelligenceFraud And ImpersonationPlatform Safety And ContentFinancial ServicesProvenance And Labeling
Online Safety Act 2023, Part 5, especially sections 80–82 and section 230.
Law — active.
CitationNot recorded in registry
Covered actors
Providers of in-scope internet services publishing or displaying regulated provider pornography. The duty can apply to a service even where the provider is not UK-established.
Neutral effect or obligation summary
Use age verification, age estimation, or both, in a way that is highly effective at correctly determining whether a particular user is a child. Ensure children are not normally able to encounter regulated provider pornography. Keep an understandable written record of the age-assurance methods and how privacy was considered. Publish a summary of the record. A self-declaration of age, without more, is not age verification or age estimation under section 230. Ofcom’s 2026 report states that it does not regard age inference as highly effective for the relevant porn/minimum-age duties on the evidence described in the report.
DatesNot recorded in registry
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
BiometricsPrivacy And Data ProtectionPlatform Safety And ContentChild Safety
Online Safety Act 2023, sections 66–69; Online Safety (CSEA Content Reporting by Regulated User-to-User Service Providers) Regulations 2026, S.I. 2026/268.
Law — active.
CitationNot recorded in registry
Covered actors
Providers of regulated user-to-user services, including UK and non-UK providers meeting the OSA scope test. The reporting regulations allow for a provider or qualifying third-party provider to submit reports in the specified manner.
Neutral effect or obligation summary
Operate systems and processes that secure, so far as possible, the reporting of detected and unreported CSEA content to the NCA. Register with the NCA reporting portal before reporting where required by the regulations. Submit reports in the required format, with required available information, within urgency-based time frames. Retain and handle information as required by the reporting regulations. Section 66 applies only to CSEA content detected on or after the provision comes into force.
Effective or commencement
Apr 7, 2026
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
Synthetic MediaIdentity And KycPlatform Safety And ContentChild Safety
Online Safety Act 2023, sections 187–188; Sexual Offences Act 2003, sections 66A–66D as inserted or amended.
Law — active.
CitationNot recorded in registry
Covered actors
Any person whose conduct satisfies the offence elements in the relevant jurisdiction. Platforms may separately be regulated under the OSA.
Neutral effect or obligation summary
Section 66A covers intentionally sending a photograph or film of a person’s genitals to another person without consent and without a reasonable belief in consent, subject to statutory elements and exemptions. Section 66B covers sharing or threatening to share an intimate photograph or film without consent and without a reasonable belief in consent, including certain situations involving intent to cause alarm, distress or humiliation, obtaining a gain, or causing loss. The statutory language covers material that “shows or appears to show” a person in an intimate state. This can encompass some manipulated or synthetic material the legal question is the statutory definition and evidence, not whether a file is labelled a deepfake. Sections 66C–66D provide exemptions and interpretation rules.
Effective or commencement
Jan 31, 2024
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
Synthetic MediaPlatform Safety And ContentProvenance And Labeling
Data (Use and Access) Act 2025, section 138; Sexual Offences Act 2003, sections 66E–66H.
Law — active.
CitationNot recorded in registry
Covered actors
Persons who intentionally create or request the creation of a purported intimate image of an adult without the subject’s consent and without a reasonable belief in consent. A request can be directed to a person or made generally available and can be committed whether or not the image is eventually created.
Neutral effect or obligation summary
Section 66E criminalises intentionally creating a purported intimate image of another person where the person shown does not consent and the creator does not reasonably believe the person consents. A “purported intimate image” appears to be, or include, an image of the person, appears to be of an adult, and appears to show the person in an intimate state. It need not be, or be only, a photograph or film of the person. Section 66E includes a defence of reasonable excuse. Section 66F criminalises intentionally requesting creation, or requesting inclusion/exclusion of a particular feature, without consent or reasonable belief in consent. It includes requests made generally and applies whether or not creation occurs. Section 66F includes a defence of reasonable excuse. New section 66H allows prosecution within three years of the offence and within six months of sufficient evidence coming to the prosecutor’s knowledge, subject to the statutory wording.
Effective or commencement
Feb 6, 2026
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
United Kingdom General Data Protection Regulation (UK GDPR), retained and amended Regulation (EU) 2016/679; Data Protection Act 2018, 2018 c. 12.
Law — active.
CitationNot recorded in registry
Covered actors
Controllers, processors, joint controllers, public bodies and organisations processing personal data, including developers, deployers, platforms, identity providers, biometric vendors and AI suppliers where their processing falls within scope.
Neutral effect or obligation summaryNot recorded in registry
DatesNot recorded in registry
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
Synthetic MediaArtificial IntelligenceIdentity And KycBiometricsFraud And ImpersonationPrivacy And Data ProtectionPlatform Safety And Content
UK GDPR Article 9; DPA 2018; ICO Biometric data guidance: Biometric recognition.
Law supplemented by regulator guidance.
CitationNot recorded in registry
Covered actors
Controllers and processors using biometric data
the analysis can apply to face, voice, physiological or behavioural data used for identification or verification.
Neutral effect or obligation summary
Biometric data is special-category data when processed for the purpose of uniquely identifying a natural person. A controller must identify both an Article 6 lawful basis and an Article 9 condition. The ICO states that explicit consent is likely to be the most appropriate condition in many biometric-recognition situations, but the correct condition is fact-specific. A controller must assess necessity, proportionality, accuracy, bias/discrimination, transparency, security, retention, data-subject rights and alternatives. The ICO states that a DPIA must be completed before using a biometric-recognition system where the processing is high risk its guidance says a DPIA is required before biometric recognition. Controllers should consider the supply chain, processors, template protection, privacy-enhancing technologies and data minimisation.
DatesNot recorded in registry
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
Data (Use and Access) Act 2025, section 92 (inserting Data Protection Act 2018 sections 124A–124B), section 80 and related automated-decision provisions; Data Protection Act 2018 (Code of Practice on Artificial Intelligence and Automated Decision-Making) Regulations 2026, S.I. 2026/425.
Law — active, requiring future statutory guidance/code preparation.
CitationNot recorded in registry
Covered actors
The ICO must prepare the code
controllers and processors are the organisations expected to use the resulting guidance in relevant AI/ADM processing.
Neutral effect or obligation summaryNot recorded in registry
Effective or commencement
2026
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
Artificial IntelligenceBiometricsPrivacy And Data Protection
Fraud Strategy 2026 to 2029: disrupting crime, supporting economic resilience and delivering justice.
Government strategy/policy; not a new offence or binding technical standard.
CitationNot recorded in registry
Covered actors
Government, law enforcement, businesses, platforms and partners referenced in the strategy
the strategy itself does not impose duties on every organisation.
Neutral effect or obligation summary
The strategy identifies generative AI, deepfakes, large language models and voice cloning as increasing fraud sophistication, credibility and scale. It describes government work with DSIT, the Alan Turing Institute and other departments on a framework for evaluating/detecting deepfake media, including fraudulent documents and synthetic audio.
DatesNot recorded in registry
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
Synthetic MediaArtificial IntelligenceIdentity And KycBiometricsFraud And ImpersonationPlatform Safety And ContentFinancial Services
Money Laundering, Terrorist Financing and Transfer of Funds (Information on the Payer) Regulations 2017, S.I. 2017/692, especially regulations 27–33.
Law — active.
CitationNot recorded in registry
Covered actors
“Relevant persons” such as financial institutions, credit/financial institutions, certain cryptoasset businesses, professional firms and other businesses specified by the Regulations.
Neutral effect or obligation summary
Apply customer due diligence when regulation 27 requires it. Identify and verify customers, understand beneficial ownership/control and apply ongoing monitoring. Use enhanced due diligence and monitoring in higher-risk situations, including certain non-face-to-face relationships without appropriate safeguards. Regulation 28(19) permits information to be regarded as from a reliable independent source where obtained through an electronic identification process that is secure from fraud and misuse and capable of providing the assurance necessary to manage and mitigate money-laundering and terrorist-financing risks. The Regulations do not mandate biometrics as the only KYC method. A firm remains responsible for a risk-based, adequate process.
DatesNot recorded in registry
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
Synthetic MediaIdentity And KycBiometricsFraud And ImpersonationFinancial Services
UK digital verification services trust framework 1.0; Using digital identities with the Money Laundering Regulations.
Statutory framework publication and government guidance. The DVS trust framework is not a universal identity law or a general requirement to use biometrics.
CitationNot recorded in registry
Covered actors
DVS providers seeking certification and registration
conformity-assessment bodies
organisations using DVS for identity or eligibility checks
MLR-regulated firms considering DVS for CDD.
Neutral effect or obligation summaryNot recorded in registry
DatesNot recorded in registry
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
FCA FG24/1, Finalised guidance on financial promotions on social media; Financial Services and Markets Act 2000 provisions referenced by the OSA.
FCA guidance and binding financial-promotion perimeter/rules.
CitationNot recorded in registry
Covered actors
Authorised firms, approved financial promotions, unauthorised persons relying on exemptions, social-media promoters and platforms with relevant OSA category status.
Neutral effect or obligation summary
Promotions must comply with applicable financial-promotion restrictions and prescribed communications/risk-warning rules. FCA guidance addresses social-media formats, prominence and presentation, including video-based content. A deepfake endorsement does not become lawful because it is synthetic or because a platform hosts it.
DatesNot recorded in registry
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
Synthetic MediaFraud And ImpersonationCybersecurityPlatform Safety And ContentFinancial Services
Network and Information Systems Regulations 2018, S.I. 2018/506.
Law — active.
CitationNot recorded in registry
Covered actors
Operators of essential services (OESs) in specified sectors and relevant digital service providers such as online marketplaces, online search engines and cloud-computing services within the statutory definitions. Banking and financial services were excluded where already covered by equivalent domestic regulation.
Neutral effect or obligation summary
Take appropriate and proportionate technical and organisational measures to manage security risks. Prevent or minimise the impact of incidents on essential/digital services. Notify the competent authority of incidents with a significant impact on continuity. RDSPs must comply with applicable security, notification, registration and information duties.
DatesNot recorded in registry
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
Cyber Security and Resilience (Network and Information Systems) Bill, introduced 2025.
Proposal — Bill, not enacted law on 2026-08-09.
CitationNot recorded in registry
Covered actorsNot recorded in registry
Neutral effect or obligation summary
Stronger security and resilience requirements, incident reporting, supply-chain and information-sharing measures, registration and supervisory reforms, and powers to update the framework. These are proposals until enacted and commenced.
DatesNot recorded in registry
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
AI developers, deployers, operators and organisations in the AI supply chain that choose to adopt it.
Neutral effect or obligation summary
Baseline principles across AI design, development, deployment, maintenance and end of life, including protection against data poisoning, model obfuscation, indirect prompt injection and other AI-specific threats.
DatesNot recorded in registry
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
Synthetic MediaArtificial IntelligenceCybersecurityElections And Political Content
Elections Act 2022, Part 6, especially sections 41–54.
Law — active, with statutory guidance.
CitationNot recorded in registry
Covered actors
Promoters of qualifying electronic material and persons on whose behalf it is published, including paid-for material and certain organic material published by candidates, parties and registered third-party campaigners.
Neutral effect or obligation summary
Do not publish qualifying electronic material unless it carries, or directly links to, the promoter’s name and address and the name and address of the person on whose behalf it is published where applicable. Imprint information must be legible/audible, accessible across devices and retained on unaltered republication as specified by section 41. Section 42 covers paid-for electronic material section 44 covers other electronic material exceptions and candidate/material-specific rules appear elsewhere in Part 6.
Effective or commencement
Nov 1, 2023
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
Synthetic MediaArtificial IntelligenceElections And Political ContentProvenance And Labeling
Online Safety Act 2023, Part 3 child-safety provisions, Part 5 age-assurance provisions and related Ofcom codes/guidance.
Law — active, supplemented by statutory codes and guidance.
CitationNot recorded in registry
Covered actors
Regulated user-to-user services, search services, combined services and certain pornography providers
Category 1/2 services may have additional duties.
Neutral effect or obligation summary
Determine whether children are likely to access the service and complete the appropriate access/risk assessments. Use proportionate systems and processes to protect children from primary-priority and other content harmful to children, including specified pornography and other statutory content categories. Apply child-safety governance, design, moderation, user-support, reporting and transparency measures required by the Act and relevant codes. Pornography services must use highly effective age assurance to prevent children from normally encountering regulated provider pornography, with privacy considered. Services remain responsible for outsourced age-assurance performance.
DatesNot recorded in registry
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
Synthetic MediaArtificial IntelligencePrivacy And Data ProtectionPlatform Safety And ContentChild Safety
Neutral effect or obligation summaryNot recorded in registry
DatesNot recorded in registry
Verified date
Evidence boundary
Direct evidence section in the cited Markdown dossier; official URLs are reproduced from that section.
Synthetic MediaArtificial IntelligenceIdentity And KycBiometricsFraud And ImpersonationPrivacy And Data ProtectionCybersecurityPlatform Safety And ContentElections And Political ContentChild SafetyProvenance And Labeling