Justice Black observed in Donaldson v. Read Magazine, Inc. that advertisements may be wholly misleading although every sentence taken separately is literally true, whether because things are omitted that should be said or because the advertisement is composed or printed so as to mislead.
Any assessment of transparency in influencer marketing driven by artificial intelligence must begin by reconstructing the premises on which current regulatory and scholarly practice rests. Transparency is not a self-evident or single-valued idea. It is an interdisciplinary term of art shaped by law, by behavioural theory, by technological design and by a normative commitment to consumer autonomy and market fairness. This part therefore sets the analytical framework within which transparency is examined, locating it at the intersection of consumer protection law, persuasion theory, information economics and technological governance. The argument pursued throughout is that transparency in this setting should be understood not as the disclosure of information alone but as a multi-level epistemic condition in which the consumer is able to identify, interpret and evaluate both the commercial and the ontological character of what is placed before them.
At the doctrinal level, transparency in advertising is implicit in the prohibition of misleading and deceptive practice under consumer protection law. The regulatory standard, as expressed in frameworks such as that of the Federal Trade Commission, asks whether a representation, omission or practice is likely to deceive a reasonable consumer acting reasonably in the circumstances.1 Disclosure on this view is remedial: it cures deception by supplying the material information that permits an informed decision. The doctrine assumes that the delivery of information is sufficient to offset deception, and it therefore adopts an information-forcing model of regulation. That model has been criticised for producing formal disclosure rather than substantive knowledge, because it presupposes rational processing of information and takes no account of cognitive limits, behavioural bias or the interpretive effect of context.2 In influencer marketing driven by artificial intelligence the doctrine is placed under greater strain still, because what is concealed extends beyond the sponsorship to the identity and authenticity of the speaker.
To address that weakness the perspective of the persuasion knowledge model must be brought in. Friestad and Wright showed that consumers develop, over time, a body of persuasion knowledge that they deploy in recognising and coping with attempts to persuade them.3 Crucially, the activation of that knowledge depends not on the existence of a disclosure alone but on contextual cues, prior belief and the mental effort available. In influencer marketing, where the commercial message is embedded in personal narrative and social relationship, persuasion knowledge may fail to activate notwithstanding disclosure. The difficulty is compounded where artificial influencers are designed to reproduce human emotion, personality and relatability, so that the boundary between a natural and a manufactured identity becomes hard to perceive. The model therefore exposes a critical weakness in disclosure frameworks: they cannot deliver transparency if they do not trigger the cognitive recognition of persuasion.
The economic analysis of information asymmetry runs alongside that psychological account, treating transparency as the correction of an imbalance of information between market participants. In digital advertising markets, influencers and platforms know far more about the content, the sponsorship arrangements and the technological manipulation involved than the consumer does. Disclosure requirements are intended to reduce that asymmetry and improve market efficiency. Classical information economics assumes, however, that once information is released the market will correct itself,4 and that assumption does not hold in a digital environment characterised by information overload, bounded rationality and mediation by intermediaries. The asymmetry in influencer marketing driven by artificial intelligence arises not only from concealed sponsorship but from concealed artificiality, because the consumer may not know that the influencer is a synthetic creation. Transparency must therefore address informational asymmetry, concerning what is being marketed, and ontological asymmetry, concerning who or what is doing the marketing. Conventional economic models address only the first.
In sociological terms transparency must also be understood by reference to trust and authenticity as social constructs. Influencer marketing operates through a perception of authenticity that is socially constructed out of relatability, consistency and a sense of personal connection. Parasocial relationships, the one-sided emotional bonds that audiences form with media personalities, contribute substantially to consumer trust and to consumer behaviour.5 Influencers generated by artificial intelligence threaten that dynamic by presenting themselves as real without any real human experience behind them, producing what may be called synthetic authenticity. Transparency here means more than the provision of information: it concerns the integrity of social meaning in communication. Where consumers do not realise that an influencer is artificial, their trust rests on a false premise, which raises normative questions of manipulation and of freedom. Transparency should accordingly be redefined as a means of preserving the integrity of authenticity in mediated interaction.
The technological dimension adds a further complication, particularly with the emergence of watermarking, metadata labelling and platform-based disclosure systems. These mechanisms mark a shift from disclosure as a human practice to transparency as a design feature of the system itself. Watermarking marks digital content so as to signify its source or ownership, in forms that may be difficult to remove or alter, and platform labels and automated disclosures are intended to make transparency uniform and to reduce reliance on the voluntary compliance of influencers. Approaches of this kind fall within the larger paradigm of code as law developed by Lessig, which holds that technological architecture can regulate behaviour as effectively as legal rules.6 Transparency on this view is not only a legal requirement but a matter of platform governance and technological design. The effectiveness of such mechanisms nonetheless depends on their visibility, on their interpretability and on their incorporation into the cognition of the user, none of which has been sufficiently studied.
At the level of policy, emerging frameworks such as the European Union’s proposed regulation on artificial intelligence indicate a growing recognition of the need for algorithmic and content-level visibility. Requirements of that kind include the disclosure of material generated by artificial intelligence and the labelling of synthetic media, particularly where they are capable of deceiving the user.7 Such policy instruments nonetheless operate alongside, rather than within, the existing regimes governing disclosure in advertising. The result is an absence of coherence between models of artificial intelligence governance and consumer protection law, and a fragmented regulatory response to a problem that is in substance a single one.
Taken together, these foundations show that transparency is not a unitary or fixed notion but a multi-dimensional condition comprising legal disclosure, cognitive recognition, economic symmetry of information, social trust and technological design. The failure of existing structures is that they separate these dimensions rather than integrate them. This article therefore adopts a multimodal and interdisciplinary conception of transparency, within which disclosure labels, watermarking and platform-level mechanisms are treated as coordinated components of a single framework rather than as alternatives to one another.
The transparency problem in influencer marketing driven by artificial intelligence cannot be treated as a single regulatory failure or a single technological difficulty. It is a layered structural problem arising from the interaction of advertising law with algorithmic technology, platform governance, consumer psychology and digital markets. Transparency here concerns not only the disclosure of sponsorship but the disclosure of artificial identity, of algorithmic generation and of data-driven targeting. The problem must therefore be disaggregated into its dimensions if the failure of instruments such as disclosure labels and watermarking is to be understood.
The regulation of influencer marketing rests principally on consumer protection and advertising law, and in particular on the prohibition of deceptive or misleading advertising. Regulators such as the Federal Trade Commission in the United States and the Advertising Standards Authority in the United Kingdom developed their frameworks with human influencers and conventional endorsement in view, not with artificial influencers or ecosystems of generated content.8
The central legal constraint is that existing models of disclosure address the commercial relationship and not the authenticity of the creator of the content. That leaves a regulatory gap: an artificial influencer may disclose that a post is sponsored without disclosing that the influencer is itself a product of artificial intelligence. This is a distinct species of deception, deception as to identity, which advertising law does not adequately reach; and the premise that disclosure eliminates consumer deception is in any event contradicted by empirical work showing that in digital advertising disclosure often fails to dispel a misleading impression.9
The ethical difficulty rests on what may be called synthetic authenticity: artificial personalities are designed to project human character, relatability and emotional expression in order to win the trust of an audience. Influencer marketing succeeds because the viewer treats the influencer as a person rather than as an advertiser, and artificial influencers exploit precisely that, manufacturing a sense of authenticity through an algorithmically designed personality.10
In ethical terms the objection is less to covert advertising than to engineered persuasion, in which artificial agents are constructed to induce conduct while appearing natural and personal. That raises questions of manipulation, of autonomy and of informed consent online. Those who write on the ethics of artificial intelligence argue that transparency is required not only for accountability but to preserve human autonomy in algorithmically mediated environments.11 Without disclosure of artificial identity and of artificial generation, consumers may be unaware that they have formed parasocial relationships with non-human agents, which brings the ethical questions of emotional manipulation and the exploitation of trust to the fore.12 The ethical dimension therefore shows that transparency is not a matter of regulatory compliance but a moral requirement directed at avoiding misleading emotional involvement.
In economic terms transparency serves to reduce information asymmetry between participants in a market. Influencer marketing is a digital market in which influencers function as a source of information for purchasing decisions. Where sponsorship or artificial identity is not disclosed, consumers are placed at an informational disadvantage, which distorts market behaviour and produces inefficiency.13
The economics of influencer marketing also generate incentives against transparency. Sponsored content that appears organic attracts higher engagement and higher conversion than advertising that announces itself,14 which creates a structural tension between the maximisation of profit and the constraints of transparency, and may make ambiguity commercially attractive to advertisers and platforms alike. Artificial influencers are economically appealing for the further reason that they eliminate the risks attaching to human influencers, among them scandal, the cost of negotiation and limits of time, and they are highly scalable as instruments of advertising.15 The economic dimension therefore shows that failures of transparency are not accidental but are built into the incentive structures of platforms and advertisers.
The technological dimension concerns the movement from disclosure by a person to transparency built into the system, through watermarking, metadata tagging and automated labelling. These techniques are a step towards what has been described as architecture-based regulation, or code as law, in which regulatory objectives are given effect by the design of the technology itself.16
Watermarking may mark content generated by artificial intelligence with a visible or invisible indication of its synthetic origin, and platform labelling may inform users automatically that content is generated or sponsored. There are nonetheless significant limitations: the design of the user interface, the availability of methods for removing a watermark, and the tendency of users to disregard the labels they are shown.17 Transparency technologies must in addition operate across platforms and across jurisdictions if they are to be useful, which is a considerable technical and regulatory problem. The technological dimension therefore shows that although watermarking and labelling may increase transparency, their effectiveness depends on design, on visibility, on standardisation and on enforcement.
The effectiveness of any transparency mechanism ultimately depends on whether the consumer perceives and acts upon the disclosure. Media psychology has shown that influencer marketing operates through parasocial interaction, the one-sided emotional attachment of audiences to media personalities.18 Such attachments reduce scepticism and strengthen persuasion, which is why influencer marketing is more effective than conventional advertising.
Parasocial relationships with artificial influencers may be more complex still, since such influencers can be optimised algorithmically for relatability and engagement. Research indicates that consumers frequently fail to notice sponsored content even where a disclosure is present, particularly where the disclosure is subtle or buried in a hashtag.19 The failure is therefore not one of making disclosures but of their cognitive ineffectiveness. The social dimension accordingly requires transparency to be assessed by reference to consumer understanding and behavioural effect, and not merely by the presence of a disclosure.
Digital platforms dominate the ecosystems of influencer marketing, since they determine what content may be disseminated, what disclosure technologies exist and how visibility is allocated algorithmically. That makes platforms private regulators of transparency standards.20 Instagram, YouTube and TikTok provide branded content disclosure features, but these are enforced irregularly and depend largely on self-reporting by influencers.
State regulation meanwhile remains fragmented across jurisdictions, with different countries taking different approaches to influencer disclosure and to transparency in artificial intelligence. The Digital Services Act and the proposed Artificial Intelligence Act of the European Union place considerable weight on the transparency of generated content, whereas other jurisdictions rely principally on consumer protection law.21 That fragmentation creates gaps in enforcement and opportunities for regulatory arbitrage, and makes global transparency difficult to secure. The governance dimension therefore shows that transparency is co-produced by platform governance and state regulation and cannot be secured by either alone.
Transparency is closely connected with consumer rights, and in particular with the right to information and the right to make an independent choice. Consumers cannot choose in an informed way if they are not told that material is sponsored or that it has been generated by artificial intelligence. Informational self-determination requires transparency, because it presupposes that the individual can regulate and comprehend the information that shapes their decisions.22
On that footing transparency is not merely a regulatory requirement but a foundational consumer right in digital markets. A failure to disclose artificial identity or algorithmic targeting may therefore be characterised not only as misleading advertising but as an infringement of informational autonomy.
The allocation of liability for a failure of disclosure is among the most difficult aspects of the problem. In traditional advertising, legal responsibility rests on the advertiser and the endorser. In advertising driven by artificial intelligence the candidates for responsibility include the developer of the system, the platform, the advertiser and the supplier of the data.23
The result is a diffusion of responsibility in which every actor contributes to the production and dissemination of content and none is wholly accountable for a failure of transparency. Existing liability frameworks are ill-suited to algorithmic ecosystems because most of them presuppose human actors and direct models of control.24 Without a clear allocation of liability, transparency obligations lose their value, since they cannot be enforced.
Influencer marketing driven by artificial intelligence frequently rests on data analytics and algorithmic targeting in order to personalise content to the individual user. That introduces a privacy dimension, because consumers may not know that the influence to which they are exposed has been tailored to their own personal data, browsing history or preferences.25
Opaque algorithmic targeting produces what has been described as invisible influence, in which a person is shaped without knowing why particular content is placed before them.26 This raises the question whether transparency should extend beyond the disclosure of sponsorship and of artificial generation to the disclosure of targeting and of data practices.
Influencer marketing is international while legal control remains territorial. That creates a jurisdictional problem, since influencers, advertisers, platforms and audiences may be situated in different countries under different regulatory regimes.27
In the absence of harmonised international standards for transparency in artificial intelligence and for influencer disclosure, regulatory arbitrage becomes possible, with firms establishing themselves where regulation is weakest. Effective regulation of transparency therefore requires international cooperation and standardisation rather than domestic regulation alone.
The regulatory difficulty presented by transparency in this field arises not from an absence of legal rules but from normative contradictions between law, economic incentive, technological capability and fundamental rights. Transparency regulation crosses privacy, intellectual property, freedom of expression, platform regulation and innovation policy, each governed by different legal and philosophical premises. Attempts to regulate have therefore generated tension rather than resolution. This part argues that the transparency problem is a conflict of principles rather than a regulatory vacuum, and that recognising those conflicts is a precondition of coherent policy.
Among the most serious conflicts is that between the protection of privacy and technological innovation. Influencer marketing driven by artificial intelligence depends on the extensive use of data analytics, behavioural modelling and algorithmic personalisation in order to generate and target content and to shape individual interaction with agents of influence. Such technologies increase efficiency and creative output in digital advertising markets, but they rest on the collection and manipulation of large quantities of data over which the individuals concerned frequently have little influence. Data-driven personalisation has been said to constitute a system of behavioural prediction and influence operating largely in the shadows, with consequences for privacy, autonomy and informational self-determination.28
Strict requirements of transparency and disclosure governing the use of data and the generation of content may protect privacy and consumer sovereignty, but they also raise the cost of compliance and may impede the development of marketing technology.29 The policy dilemma is that excessive regulation may slow technological advance while insufficient regulation allows intrusive data practices and manipulative marketing to flourish. The question is therefore not whether regulation is necessary but how the encouragement of innovation is to be reconciled with the protection of privacy and of consumer control.
A second conflict concerns ownership and intellectual property in content and personas generated by artificial intelligence. The datasets used to generate artificial influencers ordinarily include publicly available images, social media content and patterns of human behaviour. That raises the question whether such personas and content should be privately owned by the companies that produce them, as intellectual property, or should be treated as part of a common stock of knowledge because they are derived from publicly available material.30
Those who build virtual influencers argue that they expend resources in creating a digital personality and should therefore hold intellectual property rights in its image, character and output. Critics respond that most systems are trained on publicly available cultural and visual material, so that a generated persona is in a real sense constructed out of public information and cannot properly be privately owned.31 The tension between private ownership and the public domain complicates the regulation of artificial influencers, particularly where transparency would require disclosure of the process of generation or of the datasets used, which firms may treat as trade secrets.
Freedom of expression and commercial speech are also implicated. Creators and influencers may object that requirements of disclosure, labelling or watermarking intrude upon their expressive or creative freedom, particularly where content combines entertainment, personal expression and advertising.32
Consumer protection and the law of reputation meanwhile require that consumers not be misled and that individuals and companies not be exposed to false endorsement or manipulated information. The law therefore seeks a balance between expression and the prevention of deception. Artificial influencers complicate that balance, because it is unclear whether algorithmically generated content is expression to be protected or commercial communication to be regulated. Whether such content is best characterised as speech, as product or as service remains contested, and each characterisation yields a different regulatory outcome.33 Transparency regulation must therefore confront the conflict between protecting expression and preventing deceptive advertising.
There is a further structural conflict between the speed of technological change and the speed of legal and regulatory processes. Artificial influencers, synthetic media and automated marketing systems have grown rapidly while the legal frameworks governing advertising and consumer protection remain founded on the traditional model of endorsement.34 This produces what has been called the pacing problem, in which the law is perpetually overtaken by technological development.35
Regulators consequently extend existing advertising and disclosure rules to circumstances they were never designed to govern, which produces regulatory uncertainty, uneven enforcement and self-regulation by platforms. Technology firms meanwhile introduce transparency solutions in the form of disclosure labels and watermarking before governments have settled formal standards, which allows private firms to shape regulatory norms.36 That in turn produces a governance conflict, since technology companies come to act as de facto regulators, raising questions of accountability, legitimacy and democratic control.
A related conflict lies between platform governance and the regulatory authority of the state. Influencer marketing ecosystems are governed on social media platforms through algorithms, content policies and disclosure tools. Platforms possess the technical capacity to give effect to transparency through automated labelling, watermarking and disclosure requirements built into their architecture. They are nonetheless privately owned companies whose principal object is profit rather than the protection of consumers.37
Governments possess regulatory authority but frequently lack the technical knowledge and enforcement capacity required to govern algorithmic advertising. The result is a regulatory gap in which transparency norms are set internally by platforms and are not matched in law. This has been described as platform governance, in which private companies effectively regulate speech, advertising and transparency in digital environments.38 The conflict is therefore between private corporate ordering and public regulation, and it raises questions of accountability and democratic legitimacy.
A further conflict arises between individual autonomy, algorithmic influence and the intervention of the state. Transparency mechanisms such as disclosure labels and watermarking are intended to protect individual autonomy, on the principle that consumers must know that content is sponsored or generated. Excessive regulation of digital content may nonetheless be criticised as an assertion of governmental power and as a restriction on access to speech.39
Algorithmic systems themselves shape individual decision-making through personalisation and targeting, often in ways the user does not perceive. Individual autonomy, algorithmic influence and state regulation therefore stand in a triangular relation in which each exercises a different form of power over the informational environment. Transparency regulation must hold the balance between ensuring that users are not manipulated and avoiding excessive state control over digital communication.
The most fundamental conflict is between ethical responsibility and the profit motive. Influencer marketing and algorithmic advertising rest on engagement, attention and advertising revenue. Content that does not appear commercial performs better than advertising that plainly is, which creates an economic incentive to make disclosures subtle or ambiguous.40
Artificial influencers are economically advantageous for the further reason that they are wholly controllable, attract no scandal and can produce content without the cost of labour at any hour.41 That makes them both economical and ethically troubling where the audience does not know that the influencer is a machine rather than a person. The tension between ethical transparency and the maximisation of profit is therefore structural, and cannot be resolved by disclosure regulation alone, because the economic logic of digital advertising markets runs against the object of transparency.
Transparency in this field has been regulated unevenly across jurisdictions, platforms and regimes, reflecting differences of legal tradition, of policy priority and of technological approach. Comparative analysis shows that although several regulatory models are available, comprising disclosure-based consumer protection, algorithmic transparency and platform self-regulation, none of them resolves the structural problem. Each jurisdiction addresses a particular facet, whether deceptive advertising, data protection or transparency in artificial intelligence, without connecting those facets into a single effective system.
Influencer marketing in the United States is regulated principally through consumer protection law, and in particular by the Federal Trade Commission’s suppression of deceptive advertising. The model presupposes that transparency is achieved where consumers are made aware of commercial intent, and enforcement action has been taken against influencers and companies that failed to reveal sponsorship relationships.42,43
The model is nonetheless concerned with commercial disclosure rather than technological disclosure. It does not directly govern artificial influencers, watermarking or the transparency of generated content. Although it addresses false advertising effectively, it does not reach the newer problem of artificial identity. Disclosure-based regulation is in any event limited, because consumers frequently do not attend to disclosures or interpret them correctly, so that legal compliance does not entail transparency in practice.44 The United States model therefore secures market transparency without addressing algorithmic or identity transparency.
The European Union has taken a broader approach, embedding transparency within data protection, platform regulation and emerging governance of artificial intelligence. The General Data Protection Regulation imposes transparency requirements in respect of automated decision-making, profiling and data processing, and requires that individuals be informed where decisions affecting them are taken by automated systems.45
More significantly, the Digital Services Act and the proposed Artificial Intelligence Act impose transparency obligations on online platforms and on systems of artificial intelligence, including the labelling of generated content and the disclosure of automated decision-making.46 These instruments reach beyond advertising disclosure to algorithmic transparency and the identification of synthetic media, and are therefore directly applicable to artificial influencers and to automated marketing content.
The European approach has its own limitations. Although it imposes effective requirements, it also creates demanding compliance standards and may expose businesses and platforms to substantial regulatory liability.47 It addresses data protection and transparency in artificial intelligence in general terms rather than influencer marketing in particular, so that the regulation is broad but not specific, with consequent gaps in implementation and enforcement.
The United Kingdom offers a further model in the Advertising Standards Authority, which regulates advertising disclosure and issues guidance on influencer marketing. The Authority requires influencers to identify advertising material and has acted against those who fail to label sponsored content appropriately.48
The United Kingdom model is generally described as co-regulatory, combining self-regulation by the industry with regulatory oversight. That confers flexibility and an ability to adapt to new technology, but it depends heavily on voluntary compliance and on cooperation from platforms.49 Although the Authority has succeeded in establishing the practice of influencer disclosure, the system continues to emphasise transparency of sponsorship over transparency of artificial intelligence, watermarking and algorithmic labelling. The model is therefore effective in enforcing disclosure but does not reach the transparency of artificial influencers.
In practice the transparency of influencer marketing is regulated more often by platforms than by governments. Instagram, TikTok and YouTube provide branded content tools that allow influencers to tag sponsored material and disclose collaborations, and they are testing automated labelling and moderation systems capable of detecting sponsored or artificial material.50
This model reflects what has been described as the privatisation of digital markets, with platforms acting as regulators in setting disclosure standards, enforcing policy and controlling the visibility of content through their algorithms and architecture.51 Platform governance nonetheless presents a problem of accountability, since platforms are privately owned and profit-seeking. Their disclosure tools are moreover self-administered, depending on influencers applying them voluntarily, so that enforcement is irregular. Platform governance therefore supplies technological solutions without democratic accountability or consistency of enforcement.
The rise of virtual influencers and the growth of artificial personalities on social media expose the weakness of transparency law directly. Such influencers are managed by firms yet present as human faces on social networks and participate in brand campaigns.52 In most cases the audience cannot readily tell that the influencer is synthetic, which raises the question of identity transparency and the disclosure of synthetic media.
Most regulatory frameworks do not expressly require disclosure that an influencer is a generated creation, because transparency requirements are framed around sponsorship rather than identity. The case of the virtual influencer therefore shows that existing regulation is designed for human influencers and leaves synthetic personalities substantially ungoverned.
Historically the regulation of advertising was directed at preventing false or misleading statements in traditional media, in television, print and radio. The emergence of influencer marketing shifted the regulatory focus to the disclosure of sponsorship relationships. Generated content and algorithmic marketing have shifted it again, towards algorithmic transparency and the disclosure of synthetic media.53
That historical progression shows that transparency regulation changes in response to technological change but consistently lags behind it. Each stage of regulation addresses the problems of the preceding technological period rather than the present one, so that regulation is reactive rather than anticipatory, which is why the rules now in force do not meet marketing ecosystems built on artificial intelligence.
The comparison shows that different jurisdictions address transparency by different means: the United States through disclosure-based consumer protection, the European Union through data protection and the regulation of artificial intelligence, the United Kingdom through co-regulation of advertising standards, and the platforms through private ordering by disclosure tools and automated labelling. Each addresses one facet of the problem, whether commercial disclosure, data transparency, platform governance or technological labelling, and none integrates those facets into a coherent whole.
The difficulty is therefore not an absence of regulatory solutions but the disintegration of regulatory strategy, in which different jurisdictions govern different elements of transparency without addressing the systemic character of the problem. Comprehensive regulation would require an integrated system combining the law of advertising disclosure, requirements of transparency in artificial intelligence, obligations of platform governance and principles of consumer protection, rather than any single regulatory pattern.
The foregoing discussion shows that existing approaches, resting on the disclosure rules of advertising law, on fragmented technological measures and on self-regulation by platforms, are structurally insufficient. The weakness is not merely a gap in enforcement but the absence of a unified approach capable of addressing commercial disclosure, artificial identity, algorithmic mediation and consumer cognition together. This part therefore proposes a normative reconstruction of transparency as a multi-layered regulatory framework uniting legal doctrine, technological design, platform governance and rights-based principle. The central claim is that transparency must be reconfigured from a single act of disclosure into a system-level and multimodal form of governance.
Transparency should be redefined as continuous and multimodal rather than as a single act. Classical models assume that transparency is achieved by releasing material information at a moment in time. Where content is dynamic, algorithmically personalised and embedded in social narrative, transparency must instead operate at several layers at once: visual, textual, technological and contextual.
That model follows from the critique of mandated disclosure, which holds that the release of information does not by itself produce understanding or informed decision.54 Transparency must instead be designed to engage consumer cognition, so that the user is able to recognise and interpret both commercial intent and artificial identity, and it must integrate disclosure labels, watermarking and platform-level indicators into a single system operating across communication channels.
A cornerstone of the proposed framework is the formal recognition of a dual transparency mandate. Existing law is concerned with commercial transparency, the disclosure of sponsorship or of a material relationship. Influencer marketing driven by artificial intelligence creates a second requirement, which may be called ontological transparency: the disclosure that the influencer or the content is generated.
The framework proposes that both should be treated as legally distinct and equally enforceable. Commercial disclosure ensures that consumers are aware of persuasive intent; ontological disclosure ensures that they are aware of the nature of the speaker. That distinction is consistent with the European emphasis on labelling generated content but goes further, by enacting the requirement within advertising law rather than treating it as a separate regulatory field.55 A failure to recognise the dual requirement produces half-transparency, in which consumers know that content is sponsored but not that it was made by an artificial agent, which impairs their capacity to choose.
The framework adopts a layered approach in which responsibility for transparency is distributed among legislators, platforms, advertisers and developers. It recognises that transparency in a multi-actor digital ecosystem cannot be secured by any one participant.
At the legal level, minimum standards should be established by legislation, including mandatory disclosure of both sponsorship and generated identity, requirements of labelling, and a minimum allocation of liability. At the platform level, digital platforms should be required to build transparency mechanisms into their systems, including automated labelling, uniform concepts of disclosure and detection of undisclosed generated content. At the technical level, developers should be obliged to embed watermarking and metadata tagging in systems that generate content, so that its provenance remains traceable.
This layered solution is consistent with the idea of code as law, in which legal frameworks are supplemented by technological architecture so that transparency does not rest on voluntary adherence alone.56 By combining the layers, the framework compensates for the enforcement limits of disclosure-based systems and reduces dependence on individual actors.
A principal weakness of existing regimes is the absence of standardisation in the form of disclosures, in terminology and in technological indicators. The framework therefore proposes common requirements, comprising uniform labels for sponsored material, universal visual indicators for generated content, interoperable watermarking across systems, and consistent requirements as to the placement and availability of disclosures.
Standardisation improves both compliance and consumer understanding by reducing ambiguity and cognitive load. Partial or ambiguous disclosure is less effective because consumers may fail to recognise or decode it.57 Regulators should therefore work with platforms and industry participants to establish internationally agreed standards of transparency, as exist already in data protection and financial disclosure.
Liability must be reconsidered in light of the complexity of the ecosystem. Frameworks that place the whole burden on advertisers or influencers are inadequate where developers, platforms and data suppliers are also involved. The framework proposes a model of distributed liability with defined roles: advertisers are responsible for ensuring that sponsored material meets disclosure standards; platforms for establishing and operating transparency mechanisms; developers for embedding watermarking and traceability in generated content; and influencers, whether human or virtual, for accurate representation and compliance with disclosure standards.
That approach follows work proposing collective responsibility in systems of artificial intelligence, on the principle that responsibility should be commensurate with the distributed character of technological ecosystems.58 Under such a model, gaps in enforcement are closed and transparency obligations cannot be evaded through the fragmentation of responsibility.
The framework treats transparency as a consumer right rather than as a regulatory burden. That right comprises the right to know that content is sponsored, the right to know that content or an influencer is generated, and the right to know how content has been personalised or targeted.
A rights-based approach is consistent with broader principles of informational self-determination and autonomy online.59 It shifts the emphasis from compliance to empowerment, and ensures that transparency exists not to satisfy a legal requirement but to enable meaningful choice. Framing transparency as a right also strengthens the case for effective enforcement.
Ethical principles of design should complement the legal and technological measures, in order to address manipulative persuasion. The framework proposes that systems of artificial influence observe the following principles: that deceptive anthropomorphism be avoided; that human and generated personalities be properly distinguished; that transparency be achieved by disclosure built into the creation of the material; and that practices deliberately concealing sponsorship or artificial identity be prohibited.
These principles derive from established principles of the ethics of artificial intelligence, among them transparency, fairness and accountability, applied to the setting of influencer marketing.60 Ethical design matters particularly because legal compliance alone may not prevent practices that exploit consumer psychology.
Ownership and data governance are addressed through a limited proprietary model of generated influencers. Although a business may hold intellectual property in a generated persona, that interest must remain subject to transparency requirements, particularly where the content derives from publicly accessible data or from the work of users.
Practices of data use should likewise be made transparent, and the use of user data to generate and target influencer content should be disclosed. That is consistent with the emphasis in data protection frameworks such as the General Data Protection Regulation on transparency in automated decision-making and profiling.61 By bringing ownership and data governance within the regulation of transparency, the framework ensures that economic incentives cannot be used to erode consumer awareness.
The proposed framework moves beyond fragmented regulatory strategies by treating transparency as a multi-dimensional model of governance rather than as a legal imperative. It combines dual disclosure, layered regulation, standardised mechanisms, distributed liability, rights-based protection and principles of ethical design into a complete model for the regulation of influencer marketing driven by artificial intelligence.
The effectiveness of transparency in digital advertising rests not on the existence of disclosure rules but on the alignment of legal practice, technology, economic incentive and consumer cognition. The framework seeks to produce that alignment by making transparency a structural feature of digital markets rather than a procedure, so that consumers are not merely informed but able to act meaningfully within advertising settings mediated by artificial intelligence.
The normative framework establishes that transparency should operate as a multi-layered system of governance rather than as a single requirement of disclosure. Normative proposals are nonetheless incomplete until they are converted into operational legal norms and institutional procedures. This part therefore transforms the framework into a structural regulatory model comprising a multi-factor legal test of compliance, a layered allocation of regulatory responsibility, an institutional enforcement architecture, a registration scheme and mechanisms of international coordination. The object is to convert the principles of transparency into a workable instrument capable of guiding courts, regulators, platforms and industry.
In order to make the requirements operational, this article proposes a dual transparency legal test by which regulators and courts may determine whether content satisfies them. The test proceeds from the principle that transparency concerns both commercial intent and artificial identity, and that disclosure is to be assessed by reference to consumer understanding rather than to compliance alone. The factors are as follows.
1. Commercial disclosure. Whether the material relationship between the influencer and the advertiser is disclosed clearly, conspicuously and in terms an ordinary consumer would understand.
2. Artificial identity. Whether the content itself makes plain that the influencer or the content is generated, virtual or synthetic.
3. Visibility and prominence. Whether the disclosure is visible, unavoidable and presented where a reasonable consumer would perceive it.
4. Technological transparency. Whether watermarking, metadata tagging or automated labelling is present to identify generated content.
5. Consumer comprehension. Whether the form of the disclosure is intelligible to an ordinary consumer rather than concealed in technical language or in a hashtag.
6. Platform compliance. Whether disclosure tools and enforcement mechanisms exist on the platform where the content appears.
7. Intent and manipulation. Whether the content is structured so as deliberately to obscure sponsorship or artificial identity.
A multi-factor formulation is required because the effectiveness of a disclosure depends on context, on presentation and on consumer perception rather than on the mere provision of information.62 The test therefore shifts the enquiry from compliance with disclosure rules to transparency itself, so that the requirements are assessed by reference to actual consumer awareness rather than technical satisfaction.
The regulatory model rests on a layered structure allocating responsibility among institutional actors, in recognition that failures of transparency ordinarily arise from incoherent responsibility and ineffective regulation. The legislature lays down statutory requirements of transparency, of disclosure and of liability. Regulatory authorities issue guidance, monitor compliance and impose penalties. Digital platforms deploy disclosure and watermarking technology and detect undisclosed sponsored or generated content. Advertisers and agencies ensure that sponsored content meets the requirements. Developers integrate watermarking, traceability and identification into the systems that generate content.
The model draws on theories of polycentric governance, in which regulatory responsibility is held by several actors rather than concentrated in one.63 A structure of that kind reduces gaps in enforcement and ensures that transparency requirements operate throughout the advertising ecosystem rather than at the point of creation alone.
Effective implementation requires a structured mechanism of enforcement. The proposed architecture comprises a transparency compliance authority, whether a dedicated regulator or a department within an existing consumer protection agency, charged with supervising transparency in generated advertising; periodic audits of platform disclosure systems, labelling arrangements and compliance; algorithmic accountability reporting, under which platforms and developers publish reports describing how generated content is labelled and disclosed; a mechanism by which consumers may report undisclosed sponsored or generated content and obtain redress; and a graduated set of sanctions comprising fines, orders for the removal of content and orders for corrective disclosure.
That architecture reflects newer approaches to algorithmic governance and platform regulation, which rely not on litigation alone but on transparency reporting, audit and accountability.64 It therefore integrates administrative supervision, platform regulation and consumer complaint.
Given the growing commercialisation of generated influencers, this article proposes that commercial artificial influencers be registered. Firms engaging such influencers for commercial purposes would submit them to a regulator or platform registry, recording the identity of the person responsible for the influencer, a statement whether the influencer is wholly generated or assisted by a human, the advertising relationships and brand alliances involved, and a certification that disclosure and watermarking measures are in place.
Registration would attach to commercial influencer marketing, as it does to broadcasting and advertising, and not to non-commercial generated material or artistic expression. A scheme of that kind would strengthen accountability and reduce anonymity in commercial activity.
Transparency regulation should also include an architecture of consumer awareness ensuring that users are not merely informed but able to respond meaningfully. That comprises standardised symbols of disclosure, overlay systems of information, and accessible descriptions of generated content and of the targeting applied to it.
Meaningful consent and awareness online are attainable through layered disclosure, in which simple information is immediately apparent and fuller information is available at further layers.65 Applied here, that approach would enable consumers to identify generated or sponsored content quickly while retaining access to fuller information about the use of data, the targeting applied and the manner of generation.
Because the field is international, the framework must include mechanisms of international coordination, comprising harmonised standards of disclosure, common watermarking protocols and cooperation in enforcement between regulators.
Regulatory networks and international organisations may develop universal principles for the labelling of generated content and for transparency in influencer marketing, as they have for data protection and digital commerce. Without internationalisation, regulatory arbitrage will arise, with firms establishing themselves in jurisdictions with weaker standards. The framework must therefore be transnational and involve standard-setting institutions.
The proposed framework carries implications in the legal, economic, social, technological, privacy, human rights and industry spheres. Assessment of those implications matters because a regulatory proposal should be not merely normatively desirable but practically implementable, economically proportionate and socially acceptable.
The framework would alter the law of advertising, of platform liability and of artificial intelligence by establishing two transparency requirements, commercial disclosure and disclosure of artificial identity. It would extend existing disclosure regimes, traditionally concerned only with material relationships between advertisers and endorsers, to the disclosure of algorithmic or artificial identity, and would bring consumer protection law, data protection law and the regulation of artificial intelligence within a single field.
A principal consequence is the redistribution of responsibility among advertisers, developers and platforms. Under the layered structure, advertisers would cease to be the sole locus of liability and responsibility would be shared across the ecosystem. That is consistent with the tendency in the regulation of digital platforms to treat platforms as regulatory intermediaries rather than as neutral hosts.66 The framework therefore continues a movement from individual to ecosystem models of liability.
The addition of a multi-factor test would further give courts a structured means of assessing the adequacy of disclosure, reducing inconsistency and improving legal certainty. Courts find it difficult to determine whether a disclosure is sufficiently clear and conspicuous, and a formal test would produce more predictable enforcement. The legal implication is therefore not regulatory expansion so much as standardisation and doctrinal clarity.
The economic consequences run in both directions. Requirements of transparency, watermarking and platform compliance would increase costs for advertisers, developers and platforms. Disclosure obligations, registration schemes and audits might raise operating costs for smaller firms and independent creators, and could create barriers to entry.
Transparency regulation may nonetheless produce long-term economic benefit by raising consumer confidence and market stability. Markets marked by information asymmetry generally suffer from an absence of consumer confidence which impairs efficiency. Transparency reduces that asymmetry and so improves both market efficiency and consumer welfare,67 which suggests that transparency regulation is not to be regarded as a cost of compliance alone.
Registration and licensing may in addition create new economic activity, in the management of artificial influencers, in transparency auditing and in the supply of disclosure technology. The framework may therefore reorganise market structure rather than merely constrain it.
The social consequences are significant, because influencer marketing shapes culture, identity, patterns of consumption and social expectation. Artificial influencers blur the line between real and virtual personality, with consequences for social attitudes, ideals of appearance, consumption and political opinion.
The framework would help to preserve informational autonomy, the capacity to decide without distorted or manipulative information. Without transparency, consumers may not realise that they have formed emotional or parasocial relationships with generated personalities, which raises ethical questions of manipulation and psychological persuasion. Transparency therefore serves not only consumer protection but social integrity, by making plain whether the interlocutor is human or artificial.
Transparency regulation may in addition reduce deceptive advertising and improve digital literacy, as users become more conscious of generated content and of algorithmic persuasion. Over time that may spread understanding of these technologies more widely, which matters as algorithms take a larger part in public communication.
A common objection to regulation is that it discourages innovation. The framework proposed here is nonetheless technology-neutral and compatible with innovation, because it regulates outcomes, namely transparency and disclosure, rather than particular technologies.
Transparency regulation may indeed promote innovation in watermarking, content authentication, labelling and algorithmic transparency, since regulatory requirements generate demand for compliance technologies. Mandatory watermarking, for example, would encourage the development of standardised systems of content identification and digital provenance.
The framework further encourages responsible innovation, aligning technological development with ethical and legal norms. Rather than restricting the development of artificial influencers, it would establish an environment in which innovation can proceed without deception, manipulation or harm. Its technological implication is therefore not the stifling of innovation but its redirection towards technologies that improve transparency.
Privacy is directly engaged, since influencer marketing driven by artificial intelligence ordinarily rests on consumer data, behavioural targeting and algorithmic profiling. Transparency requirements may oblige platforms and advertisers to disclose not only sponsorship relationships and artificial identity but the operation of targeting algorithms.
That produces a tension between transparency and privacy, since unduly strict disclosure obligations might expose algorithmic processes or patterns in the use of personal data. The framework must therefore hold a balance between transparency and the principles of data protection, among them data minimisation, purpose limitation and consent.
Transparency regulation may nonetheless strengthen privacy protection, by disclosing practices of targeted advertising and algorithmic personalisation and so enabling users to understand how their data are used. Transparency in this field may thus operate as an indirect protection of privacy by raising awareness of data-driven advertising.
Several rights are engaged, among them consumer rights, freedom of expression, privacy, autonomy and the right to information. Transparency regulation must therefore avoid becoming over-regulation capable of restricting creative expression or the artistic use of generated personalities.
Undisclosed generated advertising may at the same time infringe consumer protection and informational autonomy, because a person may be persuaded by artificial personalities without realising it. The framework therefore upholds the right to informed decision-making, which is increasingly recognised as part of digital and consumer rights in an information society.
Transparency regulation may further guard against manipulative generated content, falsified endorsement and misleading advertising, and so contribute to digital dignity and informational self-determination. The human rights implication is accordingly the protection of individual autonomy and of informational privacy in algorithmic settings.
For the advertising and influencer marketing industry the framework would alter working practice considerably. Agencies, advertisers and platforms would adopt standards of disclosure, systems for labelling generated content and mechanisms for monitoring compliance. That would professionalise the industry by introducing standardised regulatory practice of the kind found in traditional advertising.
Developers of artificial influencers would be required to build watermarking, metadata tagging and identity disclosure into their systems, and platforms would be required to provide disclosure features and detection of undisclosed sponsored or generated content.
Although compliance would impose short-term cost, it would bring long-term benefit in industry credibility, consumer trust and regulatory certainty, each of which matters to sustained growth. Industries dependent on consumer trust, among them advertising, media and digital content, are generally responsive to transparency regulation, because trust is itself a market good. The industry implication is therefore standardisation, professionalisation and stability rather than mere regulatory burden.
The analysis shows that the proposed framework would produce system-wide change in legal systems, digital markets, technological development, consumer protection, the governance of privacy and the regulation of platforms. Although it carries costs of compliance and of regulation, it also carries long-term benefits in greater consumer trust, reduced deception, improved market efficiency and responsible innovation.
The framework is accordingly institutionally viable, economically reasonable, technologically adaptable and normatively defensible, which supports the contention that transparency in this field should be developed into a complete regulatory ecosystem rather than left to disjointed rules of disclosure.
This study began from the proposition that existing legal and regulatory frameworks governing transparency in influencer marketing are inadequate to ecosystems built on artificial intelligence, principally because they treat commercial and technological disclosure as two separate regulatory problems. Conventional influencer regulation rests on the disclosure of sponsorship and on consumer protection, while the newer regulation of artificial intelligence rests on algorithmic transparency and the labelling of synthetic media. Influencer marketing driven by artificial intelligence lies at the intersection of advertising law, consumer protection, platform regulation and the regulation of artificial intelligence, and the absence of a framework addressing both transparency of commercial intent and transparency of artificial identity produces regulatory gaps, inconsistent enforcement and a risk of consumer deception. The problem is therefore not an absence of disclosure rules but an absence of an integrated framework capable of governing hybrid commercial communication.
Earlier attempts have failed largely because they rest on single-layered models of transparency, whether a hashtag, a textual disclosure or a platform-specific label. Such approaches assume that transparency follows from disclosure, whereas empirical and legal work on disclosure regulation shows that disclosure alone does not produce consumer awareness or informed decision. Mandated disclosure regimes fail through information overload, obscurity, technicality and the limits of consumer attention. Existing frameworks moreover place the responsibility for disclosure principally on influencers or advertisers, disregarding the contribution of platforms and developers. Existing strategies are therefore fragmented, actor-focused and technologically outdated, and cannot govern generated influencers, who do not fit the classical categories of endorser, advertiser or publisher. Those weaknesses are structural rather than procedural, arising from regulatory fragmentation, limited models of disclosure and the absence of governance at the level of the ecosystem.
This study has proposed a dual transparency and layered regulatory framework which reconceives transparency as a multi-dimensional regulatory objective rather than an obligation to disclose. The model unites commercial disclosure, disclosure of artificial identity, watermarking at the technological level, transparency mechanisms at the platform level, registration and an institutional architecture of enforcement. It offers a doctrinal instrument in the dual transparency legal test, by which courts and regulators may assess compliance. It offers a decentralised model of governance by distributing responsibility among advertisers, platforms, developers and regulators, and so addresses the enforcement gaps in existing frameworks. It further incorporates technological measures into legal regulation, through watermarking, automated labelling and transparency reporting, so that transparency is embedded in digital infrastructure rather than left to selective compliance.
The significance of the framework lies in its movement from disclosure regulation to transparency governance. Disclosure regulation asks whether information has been provided; transparency governance asks whether it is visible, intelligible, enforceable and embedded in the systems of communication. That shift is necessary because generated influencers represent a new form of persuasive communication in which identity itself is artificial and persuasion may be embedded in algorithmically constructed personalities and images. In such an environment transparency concerns not only the message but the source, the method and the intent of the communication.
More broadly, this study contributes to the developing character of digital regulation by showing that the regulation of artificial intelligence, advertising law and platform regulation cannot be treated as separate fields in the context of algorithmic communication. Influencer marketing generated by artificial intelligence is one instance of a larger change in which persuasion, identity, automation and commerce intersect. Environments of that kind require integrated regulatory frameworks rather than independent legal reforms, and the framework proposed here therefore extends beyond influencer marketing to synthetic media, virtual advertising, generated journalism and automated political communication.
At a more fundamental level the study raises a question about the future of communication regulation: whether the law should regulate speech by reference to content and intent alone, or whether it should also have regard to the nature of the speaker where the speaker is an artificial agent. Generated influencers destabilise the legal assumptions that communication is of human origin, that an endorsement expresses a personal opinion, and that liability is attributable to an identified person. As artificial agents take a greater part in commercial and social communication, legal frameworks will have to develop new principles governing artificial identity, algorithmic persuasion and automated influence. Transparency regulation may prove to be among the foundational principles of that law.
The central difficulty presented by influencer marketing driven by artificial intelligence is therefore not misleading advertising as such but the informational asymmetry produced by artificial identity and algorithmic persuasion. The answer is not merely to tighten the rules of disclosure but to construct a system of transparency governance combining legal regulation, technological infrastructure, institutional enforcement and instruments of consumer awareness. In proposing such a framework this study addresses the problem with which it began and contributes to the regulatory theory of commercial communication mediated by artificial intelligence.
The movement from human to generated influencers is not only a change of technology but a change in the structure of persuasive communication and of digital markets. The law must now redefine its object, from the regulation of advertisements to the regulation of systems of algorithmic influence. Transparency, understood in that way, is not merely a mechanism of consumer protection but a condition of confidence, independence and equality in an increasingly automated communicative space.
*****
1. Federal Trade Commission, Policy Statement on Deception (1983).
2. Omri Ben-Shahar & Carl E. Schneider, More Than You Wanted to Know: The Failure of Mandated Disclosure (Princeton Univ. Press 2014).
3. Marian Friestad & Peter Wright, The Persuasion Knowledge Model: How People Cope with Persuasion Attempts, 21 J. Consumer Res. 1 (1994).
4. George A. Akerlof, The Market for “Lemons”: Quality Uncertainty and the Market Mechanism, 84 Q.J. Econ. 488 (1970).
5. Donald Horton & R. Richard Wohl, Mass Communication and Para-Social Interaction: Observations on Intimacy at a Distance, 19 Psychiatry 215 (1956).
6. Lawrence Lessig, Code and Other Laws of Cyberspace (Basic Books 1999).
7. European Commission, Proposal for a Regulation Laying Down Harmonised Rules on Artificial Intelligence, COM (2021) 206 final.
8. Federal Trade Commission, Disclosures 101 for Social Media Influencers (2019).
9. Rebecca Tushnet, Running the Gamut from A to B: Federal Trademark and False Advertising Law, 159 U. Pa. L. Rev. 1305 (2011); Jacob Jacoby & George J. Szybillo, Why Disclosures in Advertisements Fail, 6 J. Pub. Pol’y & Mktg. 107 (1987).
10. Emily Hund, The Influencer Industry: The Quest for Authenticity on Social Media (Princeton Univ. Press 2023).
11. Luciano Floridi et al., AI4People: An Ethical Framework for a Good AI Society, 28 Minds & Machines 689 (2018).
12. Horton & Wohl, supra note 5.
13. Akerlof, supra note 4.
14. Nathaniel J. Evans, Joe Phua, Jay Lim & Hyoyeun Jun, Disclosing Instagram Influencer Advertising, 17 J. Interactive Advert. 138 (2017).
15. Crystal Abidin, Internet Celebrity and Influencer Culture in Digital Media, in A Companion to New Media Dynamics (2018).
16. Lessig, supra note 6.
17. Hany Farid, Digital Forensics and Image Authentication, 49 Sci. Am. 66 (2008).
18. Riva Tukachinsky, Parasocial Relationships, 21 J. Comm. 305 (2011).
19. Sophie C. Boerman, Lotte M. Willemsen & Eva P. Van Der Aa, ”This Post Is Sponsored”, 38 J. Interactive Mktg. 82 (2017).
20. Tarleton Gillespie, Custodians of the Internet (Yale Univ. Press 2018).
21. Regulation (EU) 2022/2065 (Digital Services Act).
22. Julie E. Cohen, Configuring the Networked Self (2012).
23. Woodrow Barfield & Ugo Pagallo, Research Handbook on the Law of Artificial Intelligence (2018).
24. Frank Pasquale, The Black Box Society (2015).
25. Shoshana Zuboff, The Age of Surveillance Capitalism (2019).
26. Karen Yeung, Hypernudge: Big Data as a Mode of Regulation, 20 Info. Comm. & Soc’y 118 (2017).
27. Anupam Chander, The Electronic Silk Road (2013).
28. Zuboff, supra note 25.
29. Joshua A.T. Fairfield & Christoph Engel, Privacy as a Public Good, 65 Duke L.J. 385 (2015).
30. Pamela Samuelson, Allocating Ownership Rights in Computer-Generated Works, 47 U. Pitt. L. Rev. 1185 (1986).
31. James Boyle, The Public Domain: Enclosing the Commons of the Mind (2008).
32. Rebecca Tushnet, Commercial Speech and Influencer Marketing, 52 Loy. L.A. L. Rev. 1 (2018).
33. Tim Wu, Machine Speech, 161 U. Pa. L. Rev. 1495 (2013).
34. Barfield & Pagallo, supra note 23.
35. Gary E. Marchant, The Growing Gap Between Emerging Technologies and Legal-Ethical Oversight, 13 Santa Clara Comput. & High Tech. L.J. 395 (1997).
36. Lessig, supra note 6.
37. Gillespie, supra note 20.
38. Nicolas Suzor, Lawless: The Secret Rules That Govern Our Digital Lives (2019).
39. Jack M. Balkin, Free Speech in the Algorithmic Society, 51 U.C. Davis L. Rev. 1149 (2018).
40. Evans et al., supra note 14.
41. Hund, supra note 10.
42. Federal Trade Commission, Disclosures 101 for Social Media Influencers, supra note 8.
43. In re Lord & Taylor, LLC, F.T.C. File No. 152-3181 (2016).
44. Evans et al., supra note 14.
45. Regulation (EU) 2016/679 (General Data Protection Regulation).
46. Regulation (EU) 2022/2065, supra note 21; European Commission, supra note 7.
47. Mireille Hildebrandt, Smart Technologies and the End(s) of Law (2015).
48. Advertising Standards Authority, Influencer Advertising Guidance (U.K.).
49. Julia Black, Critical Reflections on Regulation, 27 Austl. J. Legal Phil. 1 (2002).
50. Gillespie, supra note 20.
51. Suzor, supra note 38.
52. Hund, supra note 10.
53. Pasquale, supra note 24.
54. Ben-Shahar & Schneider, supra note 2.
55. European Commission, supra note 7.
56. Lessig, supra note 6.
57. Boerman, Willemsen & Van Der Aa, supra note 19.
58. Barfield & Pagallo, supra note 23.
59. Cohen, supra note 22.
60. Floridi et al., supra note 11.
61. Regulation (EU) 2016/679, supra note 45.
62. Federal Trade Commission, Dot Com Disclosures: Information About Online Advertising (2013).
63. Elinor Ostrom, Governing the Commons (1990).
64. Pasquale, supra note 24.
65. Ben-Shahar & Schneider, supra note 2.
66. Pasquale, supra note 24.
67. Akerlof, supra note 4.