1. Introduction
Under the common law, from time to time, English judges have found it necessary to “recognize” the existence of certain pre-existing common law rights, (Chafee Jr., 1947: p. 406) 1 or even forgotten equitable interests, which can be discerned from earlier cases. (Day, 1976) 2 Although judges may claim they are able to find such common laws from a careful examination of previous cases, the exact methodology used for making such determinations is still somewhat mysterious. (Reid, 1972: p. 22) 3 Moreover, in recent years, the scope for making such determinations appears to have receded somewhat, perhaps owing to a growing political sensitivity on the part of the judiciary or simply the inherent deference given to the ever-increasing number of encroaching statutes on the books. Nonetheless, the scope for Anglosphere4 judges to create—or more properly stated, belatedly “discover”—inherent common law has never been abandoned.5
It is argued here that technological innovations associated with digital images, and their dissemination over the Internet, have given rise to a need to recognize a new common law right or legal interest, perhaps even a new generation of international human right, so as to address the harms stemming from a loss of face. Specifically, the fact-based scenarios can fall into at least three recognizable patterns of dissemination: 1) “privacy in public” image capture scenarios in which uncharacteristic behavior is displayed or unfortunate mishap befalls the image subject, 2) non-consensual pornographic and suis generis images, and 3) non-satirical images which are edited or morphed, such as deepfake imagery. Dissemination of any of these images to the public at large would likely constitute a serious affront to the interests of the image subject, leading to a possible loss of face.
2. Overview—The Nature of the Problem
Misuse of the digital image by nefarious actors is increasingly being manifest in at least three distinct fact-based scenarios. In the first scenario, dissemination of images captured in public and quasi-public places, while technically “true”, will nonetheless unfairly give rise to a risk of harm for the image subject. Such images might transgress an implied privacy interest (e.g. upskirt images) or unfairly misrepresent the nature of the image subject’s character (e.g. images of a rare drinker whilst in a state of drunken stupor). Images of public mishaps (e.g. an identifiable image subject falling down the stairs or being victimized by verbal abuse) would also fall within this category of harmful image giving rise to a possible loss of face. The harm threshold, while clearly breached, would usually not rise to the level of serious harm giving rise to protection under the traditional rubric of defamation law.
A second scenario arises with non-consensual disseminations of suis generis images. Pornographic and near-pornographic images captured in private and quasi-private places clearly fall within this category, while many jurisdictions already prohibit non-consensual disseminations of the pornographic image (generally under tort or criminal law), such laws will often fail to prevent dissemination of the near-pornographic. For example, dissemination of images of bikini-clad women, surreptitiously captured in third world go-go bars, will often lead to loss of face or even a risk of violence for the women concerned. Yet, due to the high factual (nudity) threshold set for revenge pornography law, these women will often be left without a remedy.
A third scenario of harm arises from dissemination of false images of identifiable individuals. Often, this will involve the morphing of two images e.g. woman A and woman B, creating a “new” image of a “new” person. Such techniques are often employed using deepfake technology to transpose the faces of ordinary women onto the bodies of other people such as pornographic actresses. Also falling within this category would be the use of “nudifying” technology which will typically take an image of a clothed individual and, based on an outline of their figure, create a fictional image (i.e. an extrapolation) of how they might appear if nude. While some deepfake images may be absurd (e.g. a human head stylistically attached to an animal), for third parties, where the intent of the creator is to deceive, it is becoming increasingly difficult to differentiate between the authentic and the fantasy.
Technology, such as AI, is rapidly advancing and the law has had difficulty distinguishing between the satirical and the harmful. In some sense, such innovations sit on a spectrum, only one step removed from “beautify” and the more widely used photoshop technologies. What unites all of the cases above, however, is the harm which identifiable image subjects can suffer under a loss of face, particularly when the images appear to be real. For image subjects, the truth [e.g. revenge pornography] or falsehood [e.g. deepfake pornography] of the image may have little relevance to the pain they endure.
Failure to protect against these harms can have real world consequences. Such negative effects can range from teenage mental depression and damaged relationships to loss of employment opportunities (Coleman, 2005: pp. 205-234) 6 and even altered election outcomes.7
2.1. Conceptualizing the Affected Interest—Are Psychological Injuries “Real”?
It may not be universally accepted that all—or any—of the scenarios outlined above are inducing real harms. For example, we are told that during WWII, General Patton, when visiting wounded American soldiers in the field hospital, came across a soldier suffering from PTSD who asked for permission not to return to the battlefront. However, rather than entertain such concerns, Patton famously slapped the soldier and labelled him a “coward”. (Dulcinea Staff, 2023) In effect, Patton may have been rejecting all claims of psychological injury since the perceived problems were “all in one’s head”, and hence, likely imaginary.8
The modern approach essentially rejects Patton’s view and psychological troubles are increasingly recognized as real. Moreover, anecdotal evidence suggests successive generations appear to be growing increasingly more sensitive (i.e. weaker) as their concern with outward appearance and societal acceptance rises. The quest for affirmation by strangers, social media usage and digital portrayals are all temporally, and possibly causally, linked.9 This might be manifest as everything from an increased usage and acceptance of plastic surgery, to a new obsession with receiving “likes” (i.e. dopamine hits) on social media, to even the recent increase in female teenage suicides and depression.
If it is accepted that there is value in protecting individuals from the minor mishaps and setbacks that now cause them so much suffering, then it is worth exploring possible new legal avenues for doing so. Existing defamation and criminal law, for example, will usually set too high a threshold to protect one from the typical viral humiliation video which circulates on social media. A new law or legal principle appears to be needed if we are to properly protect the weak-minded and the sensitive.
2.2. The Concept of Face
A major challenge in explaining the concept of “face” to Western audiences lies in the degree to which this philosophical and legal concept is interwoven and embedded within the fabric of law and society in the East. In order to isolate the idea of face from pride while still applying it in a meaningful way, the idea of harm is introduced as a near substitute, something with which Western legal analysts will be more familiar.
The right to an image, as outlined in this paper, is rooted in and built around protecting the ancient Oriental idea of face. Yet it is also consistent with the international covenants (e.g. Art 19 of the ICCPR “respect for the reputation of others” and the Art 12 protection of the UDHR against “attack upon [one’s] honour and reputation”).
In applying the concept of “face” to the law of the digital representation (i.e. the law of the image), at least four observations may be made:
1) “Face” is not primarily concerned with commercial misuse or non-offensive exploitation,
2) “Face” is not concerned with privacy per se (e.g. trivial issues surrounding medical privacy or the geolocation of the image capture),
3) “Face” is a mechanism for maintaining social harmony and is less concerned with individual rights than collective interests, and
4) “Face” is less concerned with image capture than image dissemination.
The right to an image is simply a manifestation of the idea that reputations are earned and individuals have a right to acquire a good image through their own good deeds or good behavior. Similarly, individuals who engage in repeated bad behavior will “deserve” the image which will invariably accrue to them. The physical image (e.g. the digital representation or painting), if it is harmful and disseminated to the wider public, should only be permitted to be disseminated if it fairly reflects the underlying reality.
2.3. The Impinged Interest
The right to an image is the right of image subjects to control against unreasonable disseminations of their image. The key word here is “unreasonable”.
While societies and civilizations may differ in how they respond to the hostile artist or anti-social photographer, all societies appear to recognize the concept of the inappropriate. Even psychopaths and sexual deviants generally know when they have crossed the line and are capturing harmful or private images. For the latter, this is even evidenced by the surreptitious image capture methods they employ. Moreover, the normalization and forced societal acceptance of bad behavior would not actually convert its nature into good behavior. At most, it would simply lead to a coarsening of society and a lowering of standards.
Having established that it can be morally wrong to disseminate—and perhaps even capture—certain images or images within certain contexts, the question then becomes: what are those images or what is the nature of those contexts? In other words, when is it unreasonable to capture and disseminate the image of another?
It could be argued, under a form of the Golden Rule, that any image dissemination which causes harm to image subjects is an unreasonable dissemination. I will refer to this as the maximalist conception of the right to an image. Such a conception effectively denies any competing interest such as truth or freedom of expression or copyright.
A more modest conceptualization, which I consider the minimalist right to an image, would distinguish between fundamental and technical truth. The fundamental truth relates to repeated or characteristic behavior, such that disseminated images which are fundamentally true do not unfairly malign: they merely express considered or consensus opinion. As such, even if these images would be harmful, they would not be unfairly or unjustly harmful: the image subject would have ultimate responsibility for how they are perceived, the cameraman or visual artist would only have penultimate responsibility.
2.4. The Competing Interests
In conceptualizing a workable version of the right to an image, two over-arching and sometimes competing goals are ever-present: 1) to minimize the impact of harm on image subjects and 2) to preserve and promote public morality.
A formulation of an unrestricted right to an image could be problematic if it became simply a veto power for image subjects over any dissemination of their image. Not only would this often become unworkable in practice (e.g. in the case of livestreaming shows) but it would also be counter-productive: wrongdoers might imagine they have license to misbehave in public without risk of censure.
As conceived here, there are at least three competing interests which could impinge on the right to an image: copyright, freedom of expression and truth. In practice, these are often overlapping interests. A few examples of disseminations involving traditional and new technologies may be illustrative.
In the case of artists, both copyright and freedom of expression claims will usually be manifest and, possibly, the artist’s subjective understanding of the truth. For example, an artist working as a political dissident could employ either traditional oil painting techniques or deepfake imagery to express his disdain for a political leader. On the assumption the jurisdiction does not recognize a public figure exception, how should we understand this act? Is the medium relevant in itself or only when it is so realistic as to amount to a deception (i.e. when the deepfake image and scenario appear to be both damaging and “real”)?
It appears that there are two possible answers here. First, if the right to an image is conceived solely as a pure human right, then the outcome of the trial itself might be unaffected by whether the harm was done through an oil painting or deepfake technology. A rights violation is a rights violation, regardless of medium. However, the degree of harm suffered—and the potential for harm is obviously greater with deepfakes, given the potential for deception—might be relevant in assessing the value of damages.
The second possible answer would be to conceive of the right to an image, at the level of municipal law, as a form of expanded tort or criminal violation. Under this scenario, actual harm suffered by the claimant would likely be a requirement, both to succeed at trial and to sustain any claim for damages. Likewise, compensation would normally, but not necessarily, be assessed based on harm suffered rather than the defendant’s ability to pay. The important point is not that deepfakes will necessarily be more damaging (i.e. harmful) than traditional oil paintings. However, the new medium has the potential to be much more harmful due to its realism.
Traditional news outlets, journalists and actors are another set of competing interests that could be negatively impacted by recognition of the right to an image.
The distinction between “fundamentally true” and “technically true” will undoubtedly prove particularly problematic when news outlets are trying to establish the character or reputation of an individual whose image they do not wish to pixelate. Here, the competing interests of truth and freedom of expression must contend with what amounts to a near privacy right, i.e. the right to an image. The Strasbourg Court has considerable experience managing this sort of problem.
In regards to deepfakes, the news outlets and crime documentaries etc. would obviously prefer simply to let AI create the imaginary or stylized recreations, rather than employing paid actors to re-enact dialogues and crime scenes. Although such scenarios would likely violate a suspect’s right to an image, it stands to reason that the rights of a convicted criminal will be fewer, in practice, than those of non-convicts. Thus, it is uncertain, but likely, that the “fundamentally true or fair” formulation would permit the staging of deepfake crime scene reenactments using the real or recognizable images of a convicted criminal but not the images of a mere suspect.
Finally, the new technology of social media has allowed for the rapid horizontal or “viral” dissemination of unusual or extreme images, often from friend to friend. These images, such as public humiliation videos, may be true or doctored and can be recognized to some degree by AI. (If the image has not been altered, it should be possible to back-trace to the initial time and place it was uploaded onto the Internet.)
Unfortunately, the virality of these images is often correlated to the extremity of the human condition. For example, a certain segment of the population appears to find satisfaction in viewing gruesome images and/or the suffering of others. The rationale for legal tolerance of this form of negative individualism tends to be couched in freedom of thought or freedom of expression arguments.
The right to an image offers a “pro-society” legal response to this problem. While those who engage in image capture and dissemination have rights, they are not the only stakeholders. Those who are image subjects also have rights. Except for the special cases where those rights have been waived (e.g. professional models or actors who have received consideration for specified images), image subjects have a natural right to control against the unreasonable dissemination of their image.
2.5. Privacy and Its Shortcomings
Recognition of the right to privacy, it was once hoped, would offer sufficient scope to protect image subjects. (Rufflo, 1987: p. 193) 10 However, privacy, as a concept, has proved surprisingly inadequate in the context of the American public square and is often not really fit for purpose.11
Privacy will also be conceived differently in different societies.12 Whereas the European conceptualization is more informational and economically rooted, the American conceptualization tends to be more property based, and, in practice, often operates more as a privilege than a right. The latter may be a consequence of the distinction between the condition of privacy and the right to privacy. For example, a rich man and a homeless man may equally share the right to privacy while the rich man lives behind high groves and the homeless man is usually visible to the public.
Depending upon the jurisdiction, the right to privacy may be over-inclusive or under-inclusive or even both. An over-inclusive formulation of the privacy interest might fail to distinguish between sensitive and non-sensitive matters. For example, medical privacy would label both the information regarding a sexually transmitted disease and information regarding common hayfever as “private” without any ranking or prioritizing as to the seriousness of the matter.
Over-inclusion thus carries the risk of trivializing the privacy interest, at least in the public mind, while under-inclusion will often lead to other interests (such as dignity and reputation) simply being subsumed into “privacy”, without serious consideration as to the actual nature of the impinged interest. This has led Professor Bloustein to argue that it is the dignity interest, rather than privacy itself, which is often the real impinged interest underlying many of the key cases. (Bloustein, 1964: pp. 962-1007) 13
2.6. The Technology Bias
Friedrich Nietzsche once stated that all things are matters of mere interpretation and “whichever interpretation prevails at a given time is a function of power and not truth”. (Turley, 2023) While this statement may be taken as a criticism of the law, or philosophy, it appears that technology, too, has its biases.
For example, facial recognition technology now has many and varied applications. It can be used as an alternative to showing one’s passport when clearing checkpoints (Anushka, 2023) and it is expected to be deployed when “unlocking” digital wallets. (Burt, 2023) Yet facial recognition technology has also been criticized for having a racial bias which is “built into” the system: the technology has had a higher error rate when assessing black faces, (Najibi, 2020) perhaps due to unconscious biases of the engineers who designed it (Fu et al., 2012) .
Likewise, current law underpinning the technology of image capture and dissemination is far from neutral. The basic underlying assumptions are all libertarian. For example, aside from a limited number of categorical exceptions (such as upskirt images and military installation images), the photographer is generally free to do as he pleases. However, unfortunate image disseminations can cause psychological harms with consequences that are often as serious as any direct physical harm. (Mascia, 2023) 14 Yet despite this risk of an ensuing harm, and due to the philosophical bias of the law, the photographer is allowed to operate without a license: anyone who can afford a camera can buy and use one. We do not follow such a relaxed policy for hunters or automobile drivers: they generally require a license and the state reserves the right to revoke this license if they cause harm.
Perhaps more problematic is the way large technology firms attempt to gather data for future resale or customer manipulation. Images may be shared voluntarily with the large technology firm (e.g. by uploading onto social media) or a firm such as Google or Amazon may attempt to acquire the images directly from the public square. The introduction of Google Street View and Google Glass resulted in considerable public backlash in certain locales and were temporarily scaled back as a result. (Treacy, 2009) 15 Amazon most famously acquires images of the public square through its doorbell security service. (Burgess, 2022) However, in the future, it may also acquire images from within the home with its robot cleaner. (Johnson, 2022) Recognition of the right to an image would not prevent any of these instances of image capture although it should curtail some of the images from being disseminated.
Although image capture through such technological innovations will necessarily impinge the privacy interest, the innovations need not unnecessarily impact the interest which the right to an image seeks to protect (i.e. if the images are not disseminated). Much like the distinction between privacy in fact and the right to privacy, a distinction can also be drawn between loss of face (or harm) and the risk of loss of face (or risk of harm). In other words, images captured non-consensually (e.g. with Google Glass) will not lead to a right to an image type infringement unless disseminated.
However, if the concept of harm is conceived in an extremely expansive way, so as to include exposure to an unwanted risk (such as accidental dissemination), then the right to an image would also be impinged by the image capture process. In the abstract, it may be difficult to compare the risk of accidental future disclosure under new versus older technologies. Salesmen in both eras would likely have claimed their products or systems were “secure”. However, insofar as some understanding of complex technology may be required to properly navigate under the new framework (e.g. by controlling access to images in the cloud) the risk of accidental disclosure seems to have risen.
2.7. Philosophical Premises
The right to an image, as set out in this paper, rests on three premises. First, that the visual manifestation of an image subject which is both recognizable (e.g. digital image, cartoon, statue etc.) and harmful would likely prove more traumatic for the image subject than a non-visual assault (e.g. spoken insult or defamatory written script). Second, that image subjects have a limited right to prohibit dissemination of unflattering yet recognizable portrayals of themselves. Third, that the right of image subjects to prohibit dissemination should not be unlimited: it may be lost voluntarily by an image subject’s repeated bad or shameful acts or behavior, (i.e. acts or behavior which affirm the fundamental truth or fairness of the negative portrayal).
If it is accepted that image subjects have a right to control against unreasonable disseminations of their image, two questions arise: 1) what is an unreasonable dissemination? and 2) how can the risk of such dissemination be mitigated against?
It is in an attempt to answer these two questions that the balance of this paper will concern itself.
2.8. Legal Rationale and Formulae: An Outline
In considering the right to an image, certain factors need to be kept in mind. In the first place, the interest may be understood as a right of individuals to control against an unfair or fundamentally untrue, harmful and recognizable, image dissemination of themselves. Thus, it is not a right belonging to states or third-party individuals or extending to animals, for example. As a human rather than property right, it is also neither transferrable nor descendible.
Second, the right to an image is not only more modest than the Franco-Romanian droit a l’image (right of image), since it does not assert a right to prohibit image capture in public places, but it is also more akin to revenge pornography than child pornography law, since it is only dissemination rather every step along the production chain which is prohibited.16
Third, the right to an image only prohibits harmful image dissemination of recognizable individuals. It does not attempt to limit all disseminations which are harmful. As a practical matter, harmful disseminations would likely be permitted if the identifying features which render the image subject recognizable were successfully obscured. (In the case of decontextualized images, such as faceless upskirt photography, it is an open question whether the concept of “recognizable” should be decided on the basis of a subjective or an objective test.)
Fourth, the right only protects against unfair or fundamentally untrue image disseminations which are harmful. It does not protect against all harmful disseminations. In conceptualizing the right to an image, we need to be cognizant of other competing interests and values which have also been recognized by the court (e.g. freedom of speech, importance of truth etc.). This context will determine what is held to be unfair or fundamentally untrue.
Fifth, as a practical matter, the right is primarily concerned with horizontal rather than vertical application. In limited circumstances, e.g. police dissemination of images of “wanted” suspects or suspected criminals, the vertical rights claim may arise, however.
Sixth, the right makes no distinction between the private acts of public and private individuals. It is a reputation-based interest, and, in the absence of a counterbalancing public policy reason, it appears that public officials are also entitled to a preservation of their dignity when acting in a private capacity.
Seventh, the right to acquire or preserve a good image is an individual right that carries with it certain implied responsibilities. The most important of these is to act responsibly and behave respectfully in the public sphere. It is not reasonable to repeatedly behave badly in public and yet still expect to enjoy the same public image as someone who behaves correctly. This is the basis for the distinction between the dissemination of the technically true and the fundamentally true image: the former might encompass a first-time drinker, while the latter would only reference a drunk.
Eighth, responsibility to enforce one’s interest for scenario one type images [see below] is primarily a matter of private law. It is difficult to envisage how, as a police matter, state authorities would determine if an individual’s identity had been rendered sufficiently obscure (e.g. by examining the pixelation settings) or whether the harm threshold had been met. Determining if the right to an image had been voluntarily forfeited through the image subject’s own repeated bad behavior would be an even more arduous task.
Ninth, life in society requires compromise. With the right to the image, many of the libertarian assumptions fundamental to Western society remain in place: there are no limits placed on image capture or editing, only on image dissemination. However, as society grows more diverse and the influence of traditional (e.g. Confucian and Islamic) ideas grow, it is reasonable to seek some accommodation with other perspectives.
Tenth, the right to an image does not directly introduce “face” as a concept into Western law. At this time in history, such a move would likely be a bridge too far. Instead, the right to an image simply protects image subjects against “harm”, rather than loss of face itself, although harm is conceptualized in dignity and reputation-based terms.
Eleventh, as a practical matter, the most difficult legal issue will be development and application of the “harm” threshold. Obviously, the high threshold set under existing law is generally too high: while defamation and criminal [e.g. revenge pornography] laws protect some victims, they do not protect enough. This is obvious, as manifest from the rising number of social media inspired mental health issues among young people, drug addiction and suicide rates etc. Yet, the harm threshold should also not be set too low, either: the law should not aspire to eliminate all insults or humor from daily life.
Twelfth, the right to an image, as outlined here, contains both elements of clarity and ambiguity. As a new generation of human right, this is to be expected. Even within traditional human rights, seemingly simple and straightforward phrases such as “the right to life”, have given rise to a surprising range of meanings and interpretations (see, for example, the post-Nuremberg War Crimes sentencing, the abortion debates and, more recently, Canada’s assisted suicide program). Likewise, it is envisioned that under municipal law, carve-outs such as a newsworthiness exception for certain images may also be created. In short, the right to an image may also come to mean different things to different people although, at its core, the dignity or reputational interest will remain: all people have a qualified right to face.
2.9. Applying the Law to the Facts
Having now defined the new legal interest, it is time to consider how it might be applied in practice. Three scenarios, so labelled, will be considered. Scenario one refers to the “privacy in public” claims, such as uncharacteristic public indecency, mishap images and affronts to dignity or reputation. Scenario two encompasses revenge pornography and “near” pornography images while scenario three encompasses deepfake and AI-generated images created for harmful, rather than satirical, intent. For ease of discussion, these will be referred to simply as scenario one, two and three images, respectively.
3. Scenario One Images
A scenario one image will often arise in circumstances factually similar to a “privacy in public” claim. Such an image, captured while individuals are in a public or quasi-public place, may be considered as a loss of face image (i.e. “harmful” image) if one or more of the following circumstances is present:
1) The image depicts the image subject undergoing a serious failure or mishap which would expose him or her to ridicule or extreme embarrassment, or
2) the image subject is captured while engaged in uncharacteristic behavior, conduct or circumstances which could expose him or her to shame, ridicule or extreme embarrassment or hostility, or
3) A sensitive area of an identifiable image subject’s physical features (e.g. “creepshot” or “voyeur” photo) is captured surreptitiously.
A possible fourth circumstance relates to self-inflicted harmful image disseminations, particularly through social media. In the case of humiliation images voluntarily shared by the image subject himself, it is an open question as to whether the act of voluntarily disseminating will negate the right to an image (effectively, whether the right should operate as only a shield, or both a sword and a shield). In the context of social media and Only Fans etc. individuals may form part of closed groups and share images within those groups, in an expectation those images will not be shared or accessed by outsiders. On the other hand, such expectations may not be entirely reasonable, particularly where the image subject shared the harmful image in expectation of consideration.
Another problematic circumstance relates to old images and modern mores. For example, during the most recent federal election in Canada, old images began to be widely disseminated of Prime Minister Justin Trudeau in “blackface”. (BBC, 2019) These images had been captured consensually at a time when such humor was not generally considered offensive.
Finally, it is worth bearing in mind, the man who walks into a public place consents to being observed by others. However, consent to being observed by those who are physically present does not necessarily mean one is consenting to being observed by all people and for all future time (i.e. having one’s image captured and disseminated). In other words, consent to image capture need not extend, even by implication, to image dissemination.
3.1. Livestreaming Public Places
Livestreaming has, at times, presented a serious theoretical and technical challenge to attempts to protect the interests of the image captured.17 Private parties on Youtube will often livestream from public locales over the Internet while news and sports TV stations engage in similar disseminations. Legally, however, it is not clear whether this type of dissemination should be classified as image capture or image dissemination or both.
This discussion is often linked to facial recognition technology, as the technology theoretically allows every face in a crowded theater to be rendered identifiable. (Karlsson et al., 2007) 18 However, technology also allows individual faces to be obscured through pixelation.
If it is accepted that individuals have a right to an image, then any harmful image of themselves which is disseminated by others should be a fundamentally true and fair representation. Lawyers may debate the meaning of “fundamentally true” and “fair”. Yet it seems to be only a matter of common sense that humiliating or traumatizing images of unlucky victims should not be disseminated to the wider public without the consent of the subjects of those images.
If such images are to be disseminated without consent then, at a minimum, the images should either be reflective of characteristic bad behavior or, if not behavior-oriented, at least be pixelated so that the identity of the victims will be obscured.19
When infringements to the right to an image occur in this context, it might be advisable to distinguish between primary and secondary image disseminations. While the primary disseminator will often lack intent to cause harm (e.g. he unexpectedly captures a mishap), those who re-upload the offending image (s), i.e. the secondary disseminators, should be held to a higher standard.
It is worth re-iterating that the right to an image can only protect against an unreasonable harmful dissemination: it cannot be relied upon to prevent all disseminations or even all harmful disseminations. However, recognition of the right to an image could fundamentally alter the ways in which social media firms are allowed to operate. One possible solution, under a maximalist interpretation of the right to an image, would be to compel pixelation of faces for livestreaming of members of the public, i.e. those for whom consent to image dissemination has not been expressly given. In contrast, under the minimalist interpretation of the right to an image, such consent would only be required where there was an intent to disseminate a harmful image.
3.2. Morally Ambiguous Images
Within the context of the scenario one image, we must also consider the morally ambiguous image. Morally ambiguous and morally dubious images point to the distinction between privacy rights and the right to an image, with the right to an image often being less inclusive or protective of image subjects. Dubious or confusing visual depictions might encompass images of entirely innocent acts, immoral behavior and even early-stage criminal preparations. Dubious images represent a wide category that could be open to more than one interpretation, or characterization, and may require knowledge of special facts in order to be deciphered correctly. Such images could include a middle-aged man talking to a teenage girl of a different race, at a bus stop, or of a man talking to a young woman in an area where prostitutes are known to work. Other examples include couples entering or leaving “short time” motels, a man talking to a transgendered individual, or famous people seen holding hands with their non-spouses.
In some of these cases, innocent explanations may exist e.g. a man may be talking to his neighbor or student, or a man new to the city may be asking strangers for directions etc. If we are to assume that the photographed behavior reveals no wrongdoing, it could be argued that dissemination of the image alone without a false or misleading counterfactual explanation attached would not in itself violate the right to an image. Any offensive interpretation would then rest upon the viewer’s own discretion or imagination, rather than at the instigation of the image disseminator. In contrast, however, at least in the ECHR context, such image disseminations might well violate the Art 8 privacy rights.20
3.3. Arguments Favoring the Right to an Image over Privacy Rights
Given the under-inclusive nature of the right to an image vis-à-vis privacy rights, at least three arguments may be advanced in favor of the right to an image. First, it can be argued that morally ambiguous images such as those discussed above should be disseminated. In contrast to the privacy-protecting libertarian position, which essentially holds that anything which is legally permitted and voluntarily agreed upon can be a positive act, Confucianism recognizes the importance and value of promoting the obvious social good. Some acts, such as public drunkenness or drug consumption, adultery or even the unwillingness of fighting age men to defend their nation should be discouraged even if the jurisdiction legally permits this behavior.
Second, if it is accepted that there is value in discouraging bad behavior, then the obvious question becomes who should do so (?) No doubt, there is value in discouraging certain behavior and the responsibility for doing so may well fall upon the community writ large. The softer mechanism of public shaming, through image dissemination and discussion by other citizens, clearly offers more scope for rehabilitation and nuance than the clumsy mechanism of the law (e.g. ASBO ordinances in the UK).
Third, the concepts of under-inclusive and over-inclusive are themselves highly normative and often conclusive. They may reveal more about the legal analyst than the underlying law itself. It is only after establishing the correct level of inclusiveness that we can determine whether the legal rule itself is under or over-inclusive. In contrast to the privacy interest, however, the right to an image is much more modest and thus unlikely to ever come to be part of the “criminal’s charter”.
4. Scenario Two Images
Scenario two images refers to sui generis images which may be captured legally but for which an expectation of privacy (i.e. non-dissemination) nonetheless exists. Pornographic and near-pornographic images captured in private and quasi-private places clearly fall within this category.
Revenge pornography has been defined as the non-consensual posting of another’s sexually explicit images to the Internet or elsewhere for the purpose of embarrassing or causing emotional harm to the subject of those images (Driscoll, 2016: p. 78, 81; Sirianni, 2015) 21 Revenge pornography—sometimes more accurately referred to as non-consensual pornography—often involves the non-consensual posting of images that were originally given to another with the implied expectation of confidentiality.22
However, while such a definition will address a large percentage of the cases referred to in the media as “revenge porn”, it will not cover all scenarios. Consider, for example, the famous actress Jennifer Lawrence who took several nude self-photographs which she uploaded onto the cloud. Her account was later “hacked” and the images uploaded onto the Internet. Since that time, Jennifer Lawrence’s hacker has been convicted and sentenced to nine months in prison. (Bradley, 2018; Ahmed, 2017) However, while American federal law criminalizes “hacking”, there is still no revenge pornography law at the federal level.
The Jennifer Lawrence scenario appears to reflect a small minority of the cases. Much more typically, the revenge pornography will initially involve either the consensual creation of a short and explicit video between two lovers or the sending of nude photographs from one party to the other (Franks & Waldman, 2019) 23. At some point later in the relationship there will be a breakdown, which leads the jilted party to seek “revenge” by sharing the photos or video with the world.
Also falling within scenario two, albeit a slight variance from the above, would involve a third party, such as a computer repair employee, accessing such images while in the course of employment. The employee(s), often hiding behind the veil of anonymity, might later upload the images onto the Internet. This was thought to be how images spread of both a famous Hong Kong actor and Hunter Biden (son of the US President).
In the American context, the core legal issues with respect to revenge pornography have traditionally been twofold: 1) whether giving consent to a partner to take an image also constitutes consent to him distributing that image and 2) whether the First Amendment protects revenge pornography (Driscoll, 2016: p. 82) .24
Although the problem may have abated somewhat in recent years, revenge pornography has been a serious problem in the United States. One survey found that ten percent of former partners threaten to post sexually explicit images after breakup and sixty percent of this cohort later follow through (Eichorn, 2013) .25 Even where there is no follow-through, many will live in fear that such exposure will lead to loss of employment or relationships.
While many jurisdictions already prohibit non-consensual disseminations of the pornographic image (generally under tort or criminal law), such laws will often fail to prevent dissemination of the near-pornographic. For example, dissemination of images of bikini-clad women, surreptitiously captured in third world go-go bars, will often lead to loss of face or even a risk of violence for the women concerned. Yet, due to the high factual (nudity) threshold set under revenge pornography law, these women will often be left without a remedy.
Also falling within scenario two would be “hacking” that could result in image capture (e.g. web cameras surreptitiously and remotely activated) or images stolen from one’s computer or smartphone etc. Here, the issue of the legality or circumstance of the image capture itself will often pale as a priority: the victim’s main concern may simply be to prevent image dissemination. As the right to an image prioritizes the rights of image subjects over copyright holders, the onus for establishing consent to disseminate would then fall upon the image disseminator.26 Under the implementing legislation, failure of a nefarious image disseminator to prove such consent could be conceived as a criminal violation or a tortious matter or both.
It is thus envisioned that the right to an image can protect against not only scenario one but also scenario two and three type disseminations. Rather than a piecemeal or enumerated list type approach (e.g. revenge pornography, hacking, disinformation etc.), a single, unifying legal principle could be invoked. From the perspective of theoretical analysis, this could be extremely useful.
4.1. Criminalizing the Image vs. Criminalizing the Malicious Act
There is broad agreement on the reasonableness of criminalizing the “original sin” i.e. the first, malicious, upload. However, the third parties involved in subsequent downloads and uploads will generally have no true knowledge of the circumstances or relationship of the primary parties. There is nothing unique to the images which will distinguish the revenge pornography image from the non-revenge pornography image. Often, the images were captured in happier times. Thus, attempts to criminalize the handling of these images by third parties, as certain States have attempted to do, will likely be found unconstitutional in the US (Driscoll, 2016: pp. 105-106) .27
At times, human behavior has been unpredictable to the point of largely circumventing the legislation. For example, in the UK, which introduced revenge pornography legislation in 2015, a jilted wife found video footage on her husband’s phone involving himself and another married woman. In anger, the wife uploaded the video, presumably with the goal of causing problems for the married rival but also with the goal of causing problems for her husband. Her act succeeded on both counts: the other woman was forced to divorce and her own husband was terminated from his employment. However, the legislation had been narrowly tailored and as she was both the uploader and a third party, her acts fell outside the scope of the jurisdiction’s revenge pornography legislation. (Buchan, 2016)
It is often factually difficult to determine who uploaded the material onto the Internet. For example, if Jane Doe sent a nude photograph of herself to John Doe, which was later found on one of the websites specializing in this genre of pornography, it is likely that the photo was uploaded: 1) by Jane Doe herself; 2) by John Doe or 3) by a hacker who accessed either Jane or John Doe’s computer or phone. In some cases, it is very difficult for the police to determine who uploaded the photo, particularly in jurisdictions which do not operate on a “real ID” system. (With regards to digital images, there is no “copy decay” to distinguish copies from originals: all digital images remain exact replicas of each other, unlike traditional photos.)
Thus, in jurisdictions without a real ID requirement for Internet access, revenge pornography law has often been hampered by factual uncertainty over who initially uploaded the image (i.e. the “original sinner”). However, under the right to an image, the general rule is that third parties would be restricted from non-consensual disseminations of harmful images. Thus, the problem should largely go away under the right to an image as, in general, no one would be willing to undertake the liability risk of uploading or re-uploading harmful images of total strangers. This outcome becomes more likely if a “real ID” system of Internet access is introduced. However, where the women are professional prostitutes it is unclear if they still have a reasonable expectation their images would remain private (Driscoll, 2016: p. 107) .
4.2. Traditional (Tortious) Remedies to Control Image Distribution
The advent of the Internet has greatly expanded the scope and ease through which tortious injuries may incur as a result of wrongful image distribution online. In the American context, there are four traditional remedies under tort law to which plaintiffs may avail themselves: appropriation, public disclosure of private facts, intentional infliction of emotional distress and false light claims.
1) Appropriation
Appropriation refers to the exploitation of attributes of the plaintiff’s identity (Prosser, 1960: p. 401) . The first case to deal with this concerned the breach of an implied contract: the photographer who took the plaintiff’s photo proceeded to put it up for sale.28 Historically, many of the cases have concerned the unauthorized use of a plaintiff’s name, picture or likeness to advertise a product (Prosser, 1960: pp. 401-402) .29 However, if the plaintiff’s photo is only used as a base, and the final version has a sufficiently minimal resemblance to the plaintiff, the defendant advertiser may escape liability.30
Some States recognize both common law and statutory causes of action for appropriation. Under the typical common law appropriation cause of action it will be necessary to prove:
a) Defendant’s use of the plaintiff’s identity;
b) Appropriation of the plaintiff’s name or likeness to the defendant’s advantage (commercial or otherwise);
c) Lack of consent; and
d) Resulting injury to the plaintiff.31
Although appropriation may be difficult to establish in typical upskirt photo scenarios, where faces are visible, even if digitally altered, courts are willing to uphold the appropriation claim.32 Success for the plaintiff will rest very much on fact-specific factors as the court in Cohen v. Herbal Concepts, Inc.,33 stated identification would rest on “the clarity of the photograph, the extent to which identifying features are visible, and the distinctiveness of those features”.34 In this case, the man had identified his nude wife and nude daughter from a photo, despite their faces not being visible.35 On the other hand, if identification were impossible then the action would fail.
2) Public disclosure of private facts
This is a cause of action based on the offensive and unwanted publicity given to such non-newsworthy facts or information which are not in the public record or not well known.36 However, this tort also requires that the plaintiff’s identity be revealed from the image itself (Calvert & Brown, 2000: p. 564) .
To prevail, the plaintiff must establish a “reasonable expectation of privacy” in the images (Beasley, 2006: p. 93) .37 Context will be determinative (as the same image could be perceived as either revenge pornography or a courtship photo) (Levendowski, 2014: p. 436) . Arguably, the concept of reasonable behavior with regards to fast evolving, and at times difficult to understand or control, technology may be somewhat fluid. For example, the District of Puerto Rico has stated that a “reasonable person does not protect his private pictures by placing them on an Internet site” even if password-protected.38
3) Intentional infliction of emotional distress
Intentional infliction of emotional distress is a cause of action which might be sustained where the images have been posted on the Internet or otherwise made available.
4) False light claim
This cause of action is similar to defamation, where the images have been manipulated to create a false impression e.g. putting the head of a celebrity on the body of another person (Calvert & Brown, 2000: p. 565) . Even if no reputational harm incurs, psychological harm to the victim can result. (Calvert & Brown, 2000: p. 565) This underscores the importance of the subjective versus objective formulation of the right to an image: adoption of a subjective test would not necessarily favor the plaintiff (as they may be particularly thick-skinned) but could act as a disincentive to the would-be defendant (since the potential plaintiff’s nature as thick-skinned or thin-skinned might be unknown).
An example of the false light tort would be the use of the plaintiff’s picture in an article with which he has no reasonable connection (such as the photo of a decent model alongside a story about “man hungry” women) (Prosser, 1960: p. 414) . Another example would be inclusion of the plaintiff’s mug shot in a rogue gallery of convicted criminals when, in fact, the plaintiff had only been arrested but never convicted of a crime (Prosser, 1960: p. 399) .
A false light action is not necessarily defamatory, although it tends to be (Prosser, 1960: p. 400) . At a minimum, it must at least be something objectionable to the ordinary man under the circumstances (i.e. the test is not subjective, is not designed to protect the hypersensitive) (Prosser, 1960: p. 400) . Prosser gives the example of a street photographer who makes a minor error when writing a story: such a plaintiff would not be entitled to recover (Prosser, 1960: p. 400) .
4.3. Drawing the Line on Individualized “Harm”
The right to an image, as a dignity-based interest, takes a somewhat distinct approach: the key question is not the tortious “did the image cause actual harm?” but rather “is the image harmful?”. The latter question can be answered objectively, without any reference to the motive of the disseminator, his knowledge or intent. Barring consent to the dissemination, the expectations of the image subject would also become irrelevant. This approach, it is submitted, saves the right to an image from the risk of excessive subjectivism which might otherwise become a consideration given the “fundamentally true” or “fair” formulation.
Of course, no discussion of the illicit image, or where the line between the lawful and the unlawful should be drawn, can take place without acknowledging the centrality of the freedom of speech discussion in international law. Article 19 of the Universal Declaration of Human Rights (UDHR) states that:
Everyone has the right to freedom of opinion and expression; this right includes freedom to hold opinions without interference and to seek, receive and impart information and ideas through any media and regardless of frontiers. (UN General Assembly, 1948)
However, this article makes no reference to pornography and while it is possible that pornography had been on the minds of the writers, there is no evidence to suggest so. Indeed, the focus appears to be on intellectual freedom i.e. ideas and information, and it is unclear if pornography has an “intellectual” component. On the other hand, the plain meaning interpretation of the words suggests that the medium itself should not be controlled and any attempt to exclude pornographic or other unpopular expressions would amount to a prior restraint. The actual distinction between pornographic and erotic or artistic expression, for example, may be more subjective, culturally specific and ambiguous than is commonly realized.
Yet, it is likely overstating the case to claim that the UDHR guarantees the right to create pornography. Freedom of belief and political expression were guaranteed. So, too, were women’s rights (see article 2) and dignity interests (see article 1) which, in some analyses, pornography would be an affront to. Moreover, this conception of freedom of speech has at times been severely circumscribed: at the time the UDHR entered into effect, no Member State had completely de-criminalized pornography.
In terms of a line of demarcation, the concept of harm as it relates to the right to an image bears little connection to the legality or illegality of the image, for example. Under the proposed test, the harmful image will almost certainly include not only the vast majority of currently illegal and tortious images but also many that have traditionally been seen as lawful.
As an example of introducing a new and lower standard in practice, the change in California’s paparazzi control law applied the lower “offensive to a reasonable person” standard, replacing the “highly offensive to a reasonable person” standard, and thereby enhanced privacy protection for the photographed.39 The right to an image, it is submitted, should adopt a similar standard in regards to dissemination.
As an example of nuance under the law, in Boring v. Google,40 the “heightened expectation of privacy” standard was considered, as the highly placed cameras peered over hedges and into a backyard, exceeding the viewing capacity of a normal pedestrian passerby. This standard was applied despite the fact backyards are not completely private: people in nearby apartments or hotels can often see in, as can overhead balloons or drones etc.41
It is the contention of this paper that much of the thinking which now dominates Western narratives over the appropriate limits to image capture and dissemination has become misguided. In particular, the importance of dignity has been lost, often swallowed up by the larger privacy claims, while individual affronts or transgressions are often downplayed, unless they can fit into a larger political narrative.
This has manifest itself at times as increasing concern over group identity and group rights at the expense of the individual, and group consciousness as group guilt or group innocence. Thus, revenge pornography may be framed as a women’s rights issue even when all of the parties involved in a film’s creation and dissemination were female. At other times this has even manifest itself as analytical confusion (e.g. “what is a woman?”). Clearly, attempts to re-draw analytical lines and formulate legal tests are political acts.
The right to an image attempts to negate some of the unfortunate tendencies by stressing the importance of dignity and reputation as a “higher form” of privacy right whilst also shifting the discussion back to individual victims and individual wrongdoers.
5. Scenario Three Images
Although the concept of revenge pornography followed here relates to real people committing real sexual acts, the term, in its most expansionist sense, is sometimes associated with deepfake pornography. For example, Levendowski describes the transplanting of an ordinary person’s head onto a sexually explicit body as an example of revenge pornography, since this imagery was non-consensual.42 However, this might also encapsulate a grey area of “spoof” photography and, not being “true” revenge pornography, is excluded for the purposes of our analysis, at least when created for obvious comedic effect. Instead, this might be better understood as deepfake pornography.
Deepfake, a portmanteau of “deep learning” and “fake”, refers to audio or visual material which has been digitally altered so as to make it appear a person is doing or saying something they have not really done or said. (Chesney & Citron, 2019) This could involve, for example, a political candidate being falsely portrayed as saying something racist, misogynistic or transphobic, so as to sabotage his election campaign.
Deepfakes erode trust, a necessary component in relationships and political discourse. (Chesney & Citron, 2019: p. 895-896) Furthermore, women and marginalized groups are disproportionately targeted for abuse, including abuse such as deepfake pornography. (Chesney & Citron, 2019: p. 896) The libertarian response, of allowing this scourge to remain unregulated, would harm the most vulnerable members of society. (Chesney & Citron, 2019: p. 896) On the other hand, some libertarians argue that any restrictions on speech will actually only favor the already powerful, given their control over regulatory processes. (Chesney & Citron, 2019: p. 897)
Despite the international nature of access to images on the Internet, there are several practical (i.e. technological and legal) difficulties that often make it impossible for authorities in one jurisdiction to effectively exercise their powers abroad. For example, remote computer investigations (e.g. US authorities accessing an American citizen’s computer abroad) might amount to a search if conducted under domestic law (Perritt, 1996: p. 109) .
Arguably, conduct should be criminalized when the cost of enforcing private rights is so high the persons harmed cannot reasonably pursue legal remedies (Perritt, 1996: p. 110) . Of course, harms to the state should also be criminalized, yet these harms can be defined over-broadly to include simple political criticism (e.g. dissident images) (Perritt, 1996: p. 110) .
The right to an image suffers from the same trilemma: image capture, image dissemination and the relevant parties may all involve three or even four jurisdictions. However, an international treaty that recognized the right to an image could also address the enforcement issue.
The failure to regulate deepfake pornography presents a serious challenge to human rights, given its close association with revenge pornography and the risk that it poses in terms of emotional distress, humiliation and reputational injury (Franks & Waldman, 2019: p. 893) . Moreover, deepfake pornography is completely at odds with a society that values the pursuit of truth and is problematic because some ideas are “unanswerable” (Franks & Waldman, 2019: pp. 894-895) .
The problems of morphed images (between two or more real persons), synthetic or computer-generated images (CGI) and child pornography are often factually confusing and overlapping. For example, the face of a child could be merged with the body of an adult or vice versa. Or, sexual cartoons may have characters with no defined age, despite their wearing school uniforms. Sometimes the CGI or cartoon will be based upon real females, or at least some of their features will be.
International treaties may, in the future, address this problem with more tightly-bound definitions of the key terms. Regrettably, however, there is still no mechanism by which individuals may directly petition a UN body when they feel their rights have been violated, nor is there an international court to adjudicate such individualized claims in the same way as the ECrtHR operates.
Recognition of the right to an image could address many of these issues, particularly where the affected individual is identifiable. Under the right to an image there is no requirement that the age of the image subject be specified or even that the photo be “real”, only that it be harmful, disseminated and the victim be recognizable. Even a CGI or cartoon drawing of an identifiable individual could be a violation of the right to an image.
In practice, deepfake revenge pornography is considered much less serious than deepfake child pornography. While true child pornography is always criminalized, revenge pornography is often considered merely tortious. Thus, while the relevant authorities may share information with another jurisdiction, they are unlikely to arrest or attempt to extradite someone who is suspected of committing an act of revenge pornography abroad (i.e. suspected of uploading abroad or suspected of uploading locally where the victim and/or the image capture itself occurred abroad).
The laws to combat non-consensual pornography (“revenge pornography”), typically prohibit image dissemination but not image capture. In contrast, there is a rare but more serious type of pornography which is not only harmful but also hideously disgusting to the general populace. This type of pornography revolves around image capture of acts which violate international norms and appear to be customarily illegal (e.g. incest, child pornography, bestiality and necrophilism). Child pornography is the most discussed of these acts although many of the laws which address child pornography cover the other acts as well.
As such, discussions of the right to an image will likely appear superfluous in such a context: given the severity of the criminal penalties involved, a photographer involved in these other acts would likely consider a discussion of a right to an image violation to be the least of his concerns. Nonetheless, the issue could arise with spoof or morphed images which might, or might not, amount to criminal offences.
Thus, the right to an image may offer a useful backstop if, for some reason, ordinary criminal law fails to protect.
6. Conclusion
There can be no doubt that a basic sense of morality does exist within mainstream society. Yet this universal value is often violated by the non-consensual image disseminator in at least three specific contexts, as outlined above. What unifies the three scenarios is that in all three there is a violation of the image subject’s right to control against harmful disseminations of his or her image.
In theory, the regional and international instruments, such as the UDHR and the ECHR, should offer adequate protection against these infringements, either under the dignity or the privacy rubric. In practice, however, the dignity conceptualization is often under-inclusive and fails to provide the same level of reputational protection as the right to an image would. In contrast, the privacy interest is often over-protective, failing to recognize the nuance or subtlety required when distinguishing between the reputation-harmful and the reputation-harmless privacy infringement. In the long run, over-protection may be almost as harmful as under-protection, insofar as the affected interest risks becoming trivialized.
In response to this quandary, in order to offer both the correct type and the correct level of protection for image subjects, this paper has laid out, in skeleton form, the right to an image. The formula developed to protect this interest is, in many ways, a compromise. For some, it will swing the balance too far away from freedom of expression and, for others, it will not swing it far enough. However, Professor Solove, and the ECrtHR, show us that it is sometimes possible to find a balance, despite the binary tendency of the law.
The “fundamentally untrue or unfair” formulation is designed to distinguish between characteristic and uncharacteristic behavior of image subjects. Only images of atypical unfortunate circumstances or uncharacteristic bad behavior would be deserving of protection. Nonetheless, some will argue that no restrictions should ever be placed on the photographer operating in the public square while others will argue that the photographer has no right to ever disseminate, or perhaps even capture, images without consent. To upset both extremes suggests perhaps a good balance has been found.
As technology such as facial recognition, digital identity and Central Bank Digital Currencies (CBDC) continue to develop, privacy will increasingly be lost, yet dignity need not be. The technology under which every face in a stadium can be recognized and broadcast does not mean that every face should be. Under circumstances where full spectrum images are taken, and the photographer disseminates them without close examination, or instantaneously, or without the mens rea of one who seeks to harm or humiliate, it is one thing. Yet, where he disseminates with the intent to harm or humiliate, it is quite another. The municipal authority may need to determine the duty of care.43
Confucius tells us that the beginning of wisdom is to call things by their true name. For many years, the law has struggled to define the vague, yet heartfelt, and widely understood interest which is triggered by the disseminations of the hostile artist or anti-social photographer. Defamation has failed when the portrayal was either technically true or clearly a work of fiction, a spoof. Likewise, a privacy infringement cannot really be claimed when images, already in the public sphere, are simply morphed. To address Confucius, the term “right to an image” has been coined to describe the affected interest. It is a term which serves to solely denote visual portrayals whose disseminations can result in loss of face for the image subjects.
NOTES
1In considering this problem, Professor Chafee puts forward the traditional argument thusly: “So, when the Wrights launched their airplane at Kittyhawk, there was already in existence a law relating to airplanes and the aviator’s right to fly over another man’s land. The judge before whom the action of trespass came would merely have to discover what this law was.” Zechariah Chafee Jr., Do Judges Make or Discover Law, Harvard University School of Law, p. 405. However, it seems his own view of the law might be called the “skyscraper theory”: some beams were erected by legislators, others by judges. Together they created the edifice.
2While there may be broad acceptance of the need for judicial activism to address lacunae in a statute’s coverage, there may be less agreement on what those gaps actually are (e.g. whether they were left intentionally). Judge Day argues that judicial legislation is not only necessary, it is also contemplated by the structure of the legal system.
3Lord Reid dismissed the idea that judges merely declare the law as a “fairy tale”. He advised judges, when creating common law, to have regard to common sense, legal principle and public policy (in that order).
4Notwithstanding the tendency to emphasize the “commonality” amongst the common law legal systems, it is worth bearing in mind that, in the background, American judges operate under a system of self-designated judicial supremacy (see Marbury v. Madison) whereas English law can be characterized as possessing legislative supremacy. Under the latter, the scope for judicial review is necessarily curtailed. It is suggested that, at the margin, this may impact legal culture and attitudes towards judicial activism.
5In the past century, this power has been used by the judiciary both to create entirely new areas of the law (e.g. modern negligence) and to redefine crimes so as expand or contract the scope of criminal wrongdoing (e.g. recognition of marital rape). For the former, see Donaghue v. Stevenson [1932] AC 562, 1932 SC (HL) 31. For the redefinition of marriage, see S v HM Advocate 1989 SLT 469 (in Scotland) and R v R [1991] 2 All ER 257 (CA), [1992] 1 AC 599 (HL) (in England and Wales). This belated recognition by the judiciary of the need to more closely align the law with existing societal values stands as a testament to the likely, eventual, acceptance of some iteration of a right to an image.
6For example, Reggie Love was enrolled at Duke University on a sports scholarship. Yet, he fell asleep drunk at a fraternity house and his “friends” drew on his face and placed their genitalia against his face, before taking a photo of this and posting it on the Internet. After the image was widely distributed, Love was suspended from the Duke sports team.
7Our Foreign Staff, “Russia using ‘deepfakes’ to sway Turkish election for Erdogan, claims frontrunner”, The Telegraph, May 12, 2023. https://www.telegraph.co.uk/world-news/2023/05/12/turkish-election-deepfake-kemal-kilicdaroglu-russia-erdogan/ (Accessed Nov 20, 2023)
8Patton’s “shock therapy” may appear to us as rather crude or perhaps even dangerous. Yet, despite the seeming unsophistication, there is some evidence that psychological injuries are self-induced or partly self-induced. For example, we may hypothesize that concern over one’s feelings or acceptance within society is negatively correlated with human happiness. In other words, the less we care what others think of us, the happier we will be.
9Tyler Durden, Dozens of States Sue Meta For Sparking Youth Mental Health Crisis, Zero Hedge, Oct 25, 2023. New York Attorney General Letitia James alleges “Meta has profited from children’s pain by intentionally designing its platforms with manipulative features that make children addicted to their platforms while lowering their self-esteem.” See https://www.zerohedge.com/political/dozens-states-sue-meta-sparking-youth-mental-health-crisis?ref=biztoc.com (Accessed Nov 20, 2023.)
10In the American context, the famous Warren and Brandeis paper, The Right to Privacy, (4 Harvard L.R. 193 (Dec. 15, 1890)) had been written specifically in response to the perceived impudent photography of nineteenth century style paparazzi on the occasion of the wedding of Warren’s daughter. The authors foresaw a future of ever more powerful cameras, due to technological advancement, causing the scope for privacy to shrink, should the law fail to intervene. However, the willingness of judges to recognize the purported common law right was limited and the right did not gain much traction in the US until the NY legislature intervened. Inspired by Warren and Brandeis, New York state has the oldest privacy legislation in the United States. Privacy rights emanate from sections 50 and 51 of the New York Civil Rights Law of 1903.
11Although all human rights may be considered somewhat reactionary, insofar as they elevate the individual interest over the collective, privacy is undoubtedly one of the most reactionary. The opacity in regards to necessary versus unnecessary secrecy is deliberate.
12In the UK, free speech rights are weaker whereas privacy rights are stronger when compared to the US.
13It should be noted that although I have used the terms “dignity”, “honor” and “reputation” more or less interchangeably, these near synonyms do differentiate themselves somewhat, with a nuance which may have evolved over time. Thus, although the term “honor” may convey a masculine concept (e.g. proper adherence to military duty), the term “dignity” often evokes a rather feminine imagery. For example, a female choice to commit suicide via overconsumption of sleeping pills might result in a “dignified” death, whereas a male choice of suicide via handgun would not. On the other hand, dying on the battlefield would be honorable, although not necessarily dignified. Thus, in describing the impinged interest discussed in this paper, it would often be no less accurate to use the more neutral term “reputation”. However, in deference to reader familiarity, I have mostly stuck simply with “dignity”.
14In the American context, gun violence and mental health are seen as major interconnected social issues. Most gun deaths in the US are actually suicides.
15As an example, reaction to the introduction of Google Street View was so hostile in Germany, the company was not only compelled to pixelate faces and license plates etc., it even agreed to erase footage of individuals who have informed authorities they do not wish to be identified.
16Image “production chain” is used here to refer to areas traditionally criminalized under child pornography law (e.g. image capture, image possession, image editing, image marketing and image distribution or dissemination).
17Historically, streaming has been abused by pedophiles in some jurisdictions through a loophole under which images might be legally viewed remotely, since the images would not be “possessed” unless downloaded.
18In fact, however, although accuracy has been increasing, it appears that facial recognition operates on probabilities rather than certainty.
19A distinction might be drawn, in some cases, between behavior and circumstances that cause loss of face. However, most harmful images would fall into one of three categories: 1) morally bad acts (e.g. unprovoked violence by image subjects), 2) unfortunate acts (e.g. falling down the stairs), and 3) unfortunate circumstances (e.g. being the victim of bird droppings). Under the restrictive or minimalist formulation of the right to an image, only those in the first category, in principle, would not be entitled to limit image disseminations.
20See, for example, the famous case of von Hannover v. Germany App no 59320/00, (2005) 40 EHRR 1. More generally, there has been an evolution over time between von Hannover (no. 1) and von Hannover (no. 3). Between von Hannover (no. 1) and von Hannover (no. 2), the German Federal Court of Justice changed its approach, to comply with the ECHR requirements, and undertook a detailed analysis of the ECrtHR jurisprudence. Under these circumstances, and in view of the margin of appreciation, the ECrtHR found in von Hannover (no. 2) that Germany had not failed to comply with its Article 8 obligations. In von Hannover (no. 3), the issue of the photograph again arose, here over the Princess’s attempt to obtain an injunction prohibiting the further publication of an image taken surreptitiously of her and her husband while on holiday. However, the ECrtHR found no violation of Article 8 as the German Courts had taken into consideration the essential criteria and the ECrtHR’s case law in balancing the competing interests. On the facts, the Court accepted that the photograph in question had contributed to a debate of general interest.
21In one survey, of those victimized, 59% had their full name included with the picture, 49% had information linking the photo (s) to the social media account, 20% had their phone number shared and 16% had their home address shared. As well, such sharing caused 82% of those surveyed “significant impairment” with family or employment relationships. In another survey, a surprisingly large number of students (i.e. more than fifty percent) expressed a desire to harm or punish their former partners by publishing their sexual photos online.
22https://www.dailydot.com/politics/federal-revenge-porn-bill. (Accessed Nov 20, 2023)
23The proliferation of smart phones, with built-in cameras, since 2010 or so has led to both a meteoric rise in “selfie” culture as well as a concomitant increase in amateur pornography. It appears that these two phenomena, while distinct, are also linked, as some individuals take self-photos in the nude. (One study has estimated that ninety percent of revenge pornography victims are women.
24Sarah E. Driscoll, Supra note 30, at 82.
25Based on a survey of 1182 American adults aged 18 - 54.
26In devising the right to an image, it is worth considering the relationship between the copyright holder and the image subject. Under copyright law, the general rule holds that the physical image is the property of its creator (e.g. photographer or painter). However, revenge pornography laws implicitly recognize that intellectual property conceptualizations are not supreme: there are higher interests, implied interests, such as dignity, which sometimes must take priority.
27Rhode Island is a case in point: criminalizing distribution of such images without the subject’s consent.
28Pollard v. Photographic Co., 40 Ch. D. 345 (1888).
29William L. Prosser, Supra note 40, at 401-402.
30See Freed v. Lowe’s, Inc., 175 Misc. 616, 24 N.Y.S. 2d 679 (Sup. Ct. 1941) where the plaintiff failed and Loftus v. Greenwich Lithographing Co., 192 App. Div. 251, 182 N.Y.S. 428 (1920) where the plaintiff succeeded. Prosser speculates the difference between success and failure lies in the extent of the resemblance.
31Hoffman v. Capital Cities/ABC, Inc., 33 F. Supp. 2d 867, 873 (C.D. Cal. 1999).
32Cohen v. Herbal Concepts, Inc., 63 N.Y. 2d 379, 384 (1984).
3363 N.Y. 2d 379, 384 (1984).
34Cohen v. Herbal Concepts, Inc., 63 N.Y. 2d 379, 384 (1984).
35Cohen v. Herbal Concepts, Inc., 63 N.Y. 2d 379, 385 (1984).
36See Shulman v. Group W Prods., Inc., 18 Cal. 4th 200, 214 (1998) and Clay Calvert; Justin Brown, Supra note 339, at 563.
37See also Restatement (Second) of Torts §§ 652D & 652B (1977).
38United States v. Gines-Perez, 214 F. Supp. 2d 205, 225 (D.P.R. 2002) (declining to find that a criminal defendant had a reasonable expectation of privacy under the Fourth Amendment with respect to images posted on a password-protected, non-public website).
39Cal. Civ. Code § 1708.8 (a) (Deering 2005).
40598 F. Supp.2d 696 (W. D. Pa. 2009). Most of the Boring claims failed, with only the trespassing claim having merit. The Third Circuit found that “[p]ublication is not an element” of intrusion upon seclusion. Furthermore, only “the most exquisitely sensitive… would suffer same or humiliation” from their property appearing on Google Street View.
41The seeds for such a conception may not be obvious although US law is generally less privacy-protecting than European law. When free speech issues are involved, a large contingent of American and America-based globalist lawyers appear eager to change that. The desire to foster international standardized norms is manifest in criticisms of US law, both implicit and explicit. An implied criticism, for example, appears to lie in the concept of “heightened expectation of privacy”. This concept implicitly rejects the hard distinction between public/private or expectations existing/expectations non-existing. By framing the idea in terms of a “heightened expectation”, a new, nuanced possibility opens up: a concept of privacy without seclusion.
42The level of such spoof and photoshopped pornography may be quite high, by one estimate accounting for twelve percent of such images. Amanda Levendowski, Supra note 52, at 424.
43It is open to the implementing legislation to provide differentiated sanctions based on the duty of care and the seriousness of the infringement. For example, a scenario one infringement could result in merely a financial sanction whereas scenario two and three infringements would more likely be deemed criminal matters.