How far can you go? Shoehorning Digital Rights into Existing Human Rights Treaties

2.1 Principles of Treaty Interpretation

Like other international treaties, the interpretation of IHRL treaties is governed by the terms of the Vienna Convention on the Law of Treaties (VCLT), which represent customary international law,^[27] and refer to text, context and object and purpose as principal interpretative techniques.^[28] The VCLT also refers to a number of additional interpretative principles and supplementary tools, including good faith,^[29] related and subsequent agreements, subsequent practice, relevant rules of international law and travaux préparatoires.^[30] Most of these additional interpretative principles and supplementary tools serve either to help choose between competing possible meanings of the treaty text, or to delineate the scope of permissible interpretation.^[31] In other words, they might support or constrain the options which the main interpretative techniques – text, context and object and purpose – offer to treaty interpreters.^[32]

At a conceptual level, one can link the different methods of treaty interpretation enumerated in the VCLT to three distinct, yet overlapping, semantic sources,^[33] which undergird the binding nature of treaties as normatively justified and socially acceptable instruments, and imbue them with specific meaning: (1) The treaty language actually agreed upon; (2) the common intention of the drafters at the time in which the treaty was formulated; and (3) the legitimate expectations of parties to the treaty that it would continuously guide their conduct and mutual relations. Treaty language signifies the scope and contents of the formal agreement of the parties articulated in the relevant text, as properly understood in light of the context and the object and purpose of the treaty.^[34] Put differently, the treaty constitutes new law – a normative point of reference – that the drafters have jointly created. Invocation of the drafters’ common intention – the second semantic source – points toward a substantive reason for the bindingness of treaties: The actual ‘meeting of minds’^[35] of the parties (ie, what they intended to agree upon). This consensual feature justifies, in a legal system that still retaining strong voluntary features, conferring on international treaties a legally binding effect.^[36] Invoking the legitimate expectations of the parties – the third sematic source – involves the social convention which underlies the motivation of states to conclude treaties in the first place and to comply, over time, with their terms, including any extra-legal moral and policy rationales which support their existence and implementation (eg, what the parties can expect to obtain from having certain aspects of their international relations governed by a treaty).^[37] Finally, legitimate expectations shape the way in which the parties would react to dynamic interpretations of the treaty (eg, they may tolerate some interpretative decisions by law-interpreters, even though they might oppose them, but might reject interpretations which they do not deem to be legitimate).^[38]

While the VCLT prioritizes language-based interpretation^[39] – which it treats as the most authentic method of establishing what the parties actually agreed upon, and as a method of interpretation that serves systemic interests of legal stability and predictability^[40] – it also gives expression, as explained below, to the other two semantic sources (common intention and legitimate expectations). In addition, one may find some support in the literature to a fourth semantic source that does not emanate directly from the will of parties – an ‘objective’ interpretation of international treaties on the basis of their ‘declared or apparent’ object and purpose,^[41] principles of justice or systemic considerations.^[42] In actuality, however, it can be presumed that the parties internalized or should have internalized, if acting in good faith, such objective elements in their formal agreement, common intention or legitimate expectations.^[43] It is therefore doubtful whether a so-called objective method of interpretation truly offers a separate source of meaning, which would allow construing a legal instrument that purports to draw its normative power from the free consent of the parties^[44] in a manner that is fully independent from their will.^[45]

Significantly, there is overlap existing between the three aforementioned semantic sources, albeit partial in scope: The treaty language reflects only imperfectly^[46] – like other complex endeavors involving the translation of wishes and ideas into words – the common intention of the drafters. This is despite the fact that treaty interpretation is sometimes presented as a process designed to retrace the path from meeting of minds to text, in order to give effect to the common intention of the drafters.^[47] The limits of reliance on treaty language are illustrated by VCLT article 32, which permits resorting to supplementary means of interpretation, not only in the face of interpretative ambiguity or obscurity, but also in the face of manifestly absurd or unreasonable results brought about by textual interpretation, which the parties could not have arguably intended to reach.

One canon of interpretation which underscores the potential tension between treaty language and the common intention of the parties is the presumption that treaty provisions should not be rendered ineffective through textual interpretation, since doing so will defeat the common intention of the drafters to generate effective treaty arrangements.^[48] In the same vein, it is presumed that the parties to the treaty did not intend to disapply important rules of international law governing their relations when concluding a treaty, unless indications to the contrary are available in the text.^[49] Still, the travaux préparatoires – documents that purport to directly capture the drafters’ common intentions or expectations^[50] – play only a limited role in treaty interpretation.^[51] This is because they are often incapable of accurately reflecting the actual meeting of minds of all parties to the treaty.^[52] One may recall, in this regard, that many treaties are binding on states that did not participate in the actual drafting process and therefore did not participate in the original meeting of minds which occurred during drafting.

Both treaty language and the common intention of the treaty drafters constitute an imperfect proxy for the third semantic source – the legitimate expectations of the parties to the treaty. Such expectations may understandably change over time, in accordance with new developments in domestic or international legal, social, political economic, cultural, environmental and technological conditions. Hence, it can be questioned whether the actual concerns that occupied the original drafters should be normatively ‘future-proofed’ – that is, whether they should dominate the interpretation of the treaty many years later.^[53] Indeed, certain supplementary tools recognized by the VCLT, such as follow-up agreements, subsequent state practice and other rules of international law, reflect the need to consider post-treaty-making developments in the treaty interpretation process. Arguably, this is because such later-in-time developments generate new legitimate expectations, which are worthy of protection.^[54]

Other manifestations of party expectations, such as public statements concerning the proper meaning of treaty provisions and positions adopted by states in connection with international litigation, may also assist law-interpreters in assessing how far they can go in pursuing the dynamic interpretation of a treaty, without encountering political backlash.^[55] Reliance on this semantic source may be particularly useful when party expectations are clear, consistent and widespread across the treaty membership; however, this is often not the case, and different states may have divergent expectations about the treaty in question. What’s more, law-interpreters are also influenced by the normative position held by non-parties to the treaty, including civil society groups and third party beneficiaries, and their expectations about how a specific treaty provision should be interpreted also carry weight (even though, as a formal matter, they are not part of the treaty interpretation process).^[56] While a treaty interpretation process that is supported by all three aforementioned semantic sources is expected to produce particularly compelling interpretative outcomes, it is quite common for courts and other law-interpreters to focus on one semantic source only – the language of the treaty.^[57]

The proposition that treaty interpretation should strive, at least in the face of textual ambiguity, to capture all semantic sources enjoys implicit support from the VCLT’s reference to the principle of good faith.^[58] The principle – and other legal principles associated with it, such as effectiveness and reasonableness – has been sometimes construed to also allude to the legitimate expectations of the parties to a treaty (which may be shaped, in turn, by broad considerations of fairness and justice).^[59] Yet, the principle of good faith can also be read in a manner that reminds law-interpreters of the need to remain bound when interpreting a treaty to its actual language and the common intention of the parties. Indeed, in his 1964 ILC Report, Waldock explained that the principle of effectiveness (which he tried and failed to enunciate in a separate paragraph of the draft VCLT) derives from the principle of good faith, and may justify going beyond the explicit language of the treaty in order to give effect to the parties’ common intention. Yet, he cautioned against relying on intention in order to reach legal outcomes too far removed from the treaty language:

If the principle of effective interpretation may be said to be implicit in the requirement of good faith, there are, it is thought, two reasons which may make it desirable to formulate it in a separate article. The first is that the principle has special significance as the basis upon which it is justifiable to imply terms in a treaty for the purpose of giving efficacy to an intention necessarily to be inferred from the express provisions of the treaty. The second is that in this sphere – the sphere of implied terms – there is a particular need to indicate the proper limits of the application of the principle if too wide a door is not to be opened to purely teleological interpretations.^[60]

In the same vein, three ICJ judges opined in 1959 that reasonableness, another principle which arguably derives from the principle of good faith, must be applied in treaty interpretation in a reasonable way^[61] – ie, without over- or under-extending the relevant treaty language – while meeting the expectations of parties that treaty provisions will have an actual effect.^[62]

2.2 Interpreting IHRL Treaties

Any limits placed on the interpretative options available to law-interpreters are looser, however, with regard to treaties that function like domestic constitutional instruments. These include the constitutive treaties of international organizations – such as the UN Charter^[63] – and IHRL instruments, which mirror domestic bills of rights to some extent. Such constitutional or constitution-like treaties are designed to confer permanent legal powers on international institutions for an indefinite period of time, and/or to place permanent side-constraints or conditions on governmental activities that take place across space and time, notwithstanding any particular political fluctuations. Given the temporally open-ended and legally ambitious nature of such treaties, they are often drafted with flexible language, containing elastic legal formulations (such as ‘arbitrary’^[64] or ‘actions necessary’^[65]) and/or abstract legal terms (eg, ‘inhuman treatment’,^[66] ‘adequate standard of living’^[67] or ‘international peace and security’),^[68] in order to allow for the adjustment of these constitutive treaties to changing realities.^[69] When such open-ended language is used, it can be reasonably assumed that the treaty drafters intended and that the parties to the treaty continue to hold legitimate expectations that the treaty in question will be construed as a ‘living instrument’.^[70]

Waldock has expressed this distinction by invoking the centrality of teleological tools when interpreting ‘constituent treaties’. Still, he cautioned about the need to respect, even in such cases, certain interpretative limits:

The point is of particular consequence in the interpretation of constituent treaties of international organizations and although those treaties, by their functional nature, may legitimately be more subject to teleological interpretations, there is evidently some limit to what may be deduced from them and still be considered ‘interpretation’.^[71]

These outer interpretative limits may be captured, to some extent, by the good faith principle, which encompasses, as discussed above, the principles of reasonableness and effectiveness, and perhaps also the expectation of the parties that their legitimate interests would be taken into consideration in the process of treaty interpretation and application.^[72] Such limits may be respected through insistence on linking the interpretative process to the three aforementioned semantic sources – treaty language, drafters’ intentions and party expectations – which may, inter alia, help to delineate the scope of permissible interpretations. In any event, the allusion by Waldock to ‘some limit’ seems to reflect a strong legal intuition that norms without definitional or conceptual boundaries have no real meaning – that is, that the interplay between inclusion and exclusion is what actually confers meaning on legal norms.^[73] In the same vein, the notional existence of definitional or conceptual boundaries – porous and difficult to pinpoint as they might be – may assist law-interpreters in evaluating the reaction of key stakeholders to the interpretative outcome: Whereas specific interpretations that reside within well-accepted definitional boundaries are unlikely to create much push-back, interpretations that challenge such boundaries are likely to encounter more resistance.^[74]

Like constitutive treaties, the interpretation of IHRL treaties often features evolutive dynamics.^[75] Indeed, IHRL treaties often mirror in their language national bills of rights.^[76] Such similarity in language may, in and of itself, signify the intention of the treaty drafters to regard IHRL treaties in the same manner in which many national jurisdictions regard their own constitutional instruments^[77] – that is, as living instruments. Furthermore, unlike treaties concluded in certain areas of international law – especially, territorial delimitation and international economic law, where a special premium is placed on legal values such as stability, predictability and mutual reliance^[78] – IHRL treaties prioritize abstract values such as justice, fairness, equality, human dignity and liberty, whose actual implementation is typically context-dependent and may very well change across time and place.^[79] This is another indication of the intention and expectation of the parties to allow for dynamic interpretation.

The longevity of IHRL treaties may create additional party expectations that the treaty would be adapted to new societal conditions and challenges through dynamic interpretation,^[80] resulting in evolution of its contents over time.^[81] Such a normative change is sometimes necessary in order to preserve the original equilibrium between competing interests introduced by the drafters.^[82] In the European Convention on Human Rights (ECHR) context, the process of periodic updating of IHRL is sometimes facilitated through treaty protocols,^[83] whereas in most of the other IHRL treaty regimes the likelihood that amending protocols would be concluded is very low. Still, even under the ECHR, a significant part of the process of normative adaptation is undertaken by evolutive interpretation techniques, which includes, inter alia, application of the margin of appreciation doctrine with its exploration of emerging European consensus on the manner in which certain Convention rights should be implemented^[84] (potentially resulting in the labelling of state conduct that was once deemed lawful by the European Court of Human Rights [ECtHR] as unlawful).^[85] This latter approach mirrors to some extent the US constitutional law ‘evolving standards of decency’ jurisprudence,^[86] and points away from following an ‘originalist’ approach when undertaking IHRL treaty interpretation.

Regarding IHRL treaties as living instruments explains why law-interpreters often resort to purposive (or teleological) interpretation, as opposed to purely textual interpretation. It also explains the preference they afford to the ‘abstract intention’ of the parties to protect the most fundamental human rights concerns relevant at the time of treaty application over the drafters’ ‘concrete intention’ to protect the human rights concerns that were considered or envisioned when the treaty was concluded.^[87] Arguably, evolutive interpretation based on abstract intentions is more consistent with promoting the humanitarian object and purpose of IHRL treaties than a drafting-era contemporaneous interpretation, which might lead to outdated results.^[88] The living instrument approach also explains the responsiveness by some treaty-interpreters to changes in general international law,^[89] and their interest in preserving through interpretation the relevant treaties’ long-term effectiveness.^[90]

Finally, the evolutive features of IHRL treaties are accentuated by the procedural opportunities presented by the continuing demands for re-interpretation encountered by IHRL bodies engaged in the periodic review of state human rights reports or the processing of large numbers of legal complaints against states.^[91] The iterative process of interaction between a variety of stakeholders – including states, civil society groups, right holders, domestic law actors, academics etc. – and law-interpreting and law-applying bodies injects additional dynamism into the process of interpreting and applying IHRL treaties.

Still, there are limits to the willingness of law-interpreters to resort to dynamic interpretation. For example, judges in IHRL courts might be averse to be seen by states and other constituencies as embarking on undemocratic ‘judicial legislation’.^[92] They might therefore be reluctant to interpret IHRL treaties in ways that directly conflict with explicit treaty language,^[93] the clear intention of the drafters^[94] and/or strong party expectations.^[95] This is because introducing new meaning into an IHRL treaty in a process that challenges the semantic boundaries set by legally and politically acceptable cannons of interpretation – ie, embracing what can be regarded as ultra-dynamic interpretation – may be viewed by some stakeholders as effectively amounting to improper treaty revision.^[96] It might also be regarded as running contrary to the legal duty placed on law-interpreters to act themselves in good faith when interpreting treaties.^[97] Such concerns about excessive judicial activism – which may manifest themselves either in significant expansion or restriction of IHRL protections^[98] – implicate the legitimacy of the treaty interpretation process and the institution facilitating it.^[99] Perhaps because of these legitimacy considerations, IHRL courts have generally tended to justify interpretative choices by invoking the legally and politically interpretation methods provided for at the VCLT (bearing in mind the particular object and purpose of IHRL treaties).^[100]

Another concern related to over-broad interpretation of IHRL treaties involves the risk of creating mega-human rights, which encompass under their purview a large number of legal norms that could be regarded as distinct human rights.^[101] Such mega-human rights might lack the degree of specificity necessary for generating effective conduct-guidance for right-holders, duty-holders and third parties. Their existence also raises difficulties implicating all three of the aforementioned semantic sources: It is doubtful whether language which appears to describe a single human right can reasonably carry on its shoulders what is effectively a cluster of different human rights, especially when read against a context in which IHRL treaties tend to introduce different rights through distinct treaty provisions. It is also questionable whether there are clear indications in the travaux préparatoires or elsewhere of the drafters’ intention or party expectation to create mega-human rights. What’s more, mega-human rights might also bring together under the same legal ‘roof’ specific norms that serve different objects and purposes^[102] – an attribute which could further complicate the treaty interpretation process.

The right to privacy appears to have already become in recent years a mega-human right, featuring multiple normative cores, and thereby illustrating the problématique of mega-human rights: Article 17 of the International Covenant on Civil and Political Rights (ICCPR) and article 8 of the ECHR, for example, are said to cover, inter alia, matters relating to preservation of unhindered intimate spaces,^[103] protection of private life in private and public spaces,^[104] limits on police search powers,^[105] protection of personal reputation^[106] and sexual and reproductive choices,^[107] measures against air pollution^[108] and forced evictions,^[109] and a swath of digital rights discussed below, including data protection and human decision-making.^[110] The normative objects and purposes underlying the different legal protections encompassed by the right to privacy also vary greatly – ranging from a right to be left alone,^[111] a right to shape one’s interactions and relationships with the broader society,^[112] allowing individuals control over their personal data,^[113] and protecting the integrity of personal decision-making and personal choices.^[114]

Such multiplicity of normative cores and objectives renders the task of interpreting the right to privacy in light of its object and purpose particularly challenging, as different right-components pull in different normative directions. Furthermore, as a practical matter, the growing breadth and complexity of the right’s scope of coverage renders the focused monitoring of its application through the work of institutions such as the Special Rapporteur on the Right to Privacy or a National Privacy Commission nearly impossible. This is especially so, given the large and diverse set of real-life issues, which are covered by this mega-human right, the complex involvement of a multiplicity of direct and indirect duty holders which should be monitored (including polluting business enterprises and technology companies who generate affordances^[115] through network effects),^[116] and the controversy surrounding the broad interpretation underlying some components of this mega-rights that appears to exceed the scope of legitimate meaning delineated by the three said semantic sources (ie, ultra-dynamic interpretation).^[117] All of these considerations raise concerns about the practical utility of developing mega-human rights.

Concerns about the legitimacy of ultra-dynamic interpretations of IHRL treaties and about the utility of developing mega-human rights explain, perhaps, the relative paucity of hard law pronouncements on the existence of digital human rights by courts and other law-interpreters, and the growing interests of norm entrepreneurs in developing digital human rights through alternatives methods to both new IHRL treaties and evolutive interpretation of existing IHRL treaties – ie, through soft law instruments, containing lex ferenda and fleshing out best practices in the field. The UN Global Digital Compact,^[118] White House AI Bill of Rights^[119] and Declaration on Digital Rights and Principles^[120] are all recent examples of instruments that elaborate new soft law norms (alongside restating some parts of existing IHRL).

3.1 Incorporation through Dynamic Interpretation

As indicated above, the emergence of new and emerging digital technologies presents a significant challenge for humanity, given the manner in which technological tools and services change the configuration of many human interactions, affect existing human needs and interests, generate completely new needs and interests which may be worthy of protection, and create new ways by which governments and other public or private actors can exercise powers of control over individuals. One response by national, regional and international lawmakers to this changing technological landscape and its potential implications for individual and societal wellbeing has been to try and create digital rights – legal rights affording entitlements and protections to affected individuals in matters relating to digital technologies. Such digital rights have been typically introduced through branches of law other than IHRL, including consumer protection, data protection and administrative law.^[121] A second response by law entrepreneurs has been to look for ways to incorporate digital rights into IHRL treaties. Given the limited appetite of states for concluding major new IHRL treaties^[122] and fears that the negotiation of a new treaty could actually result in the lowering of existing standards of IHRL protection,^[123] it is perhaps not surprising that modification of the contents of existing IHRL treaties through dynamic interpretation has been perceived by many as a central channel for incorporating digital rights into the corpus of IHRL.

The process of reading digital rights into existing IHRL treaties is informed by several related considerations: First, in some contexts, the process is relatively simple and does not pose serious problems of fit between existing IHRL and the changed technological environment; nor does it require ultra-dynamic interpretation. The most obvious example for the smooth interpretation and application of offline IHRL to the online domain is freedom of expression. This is because the freedom was formulated, for example, in article 19 of the ICCPR in a ‘future-proofed’ manner that envisions the utilization of ‘any other media’ to seek, receive and impart information.^[124] The internet, email, social media, the metaverse or text messaging all represent ‘other media’ covered by article 19, and the fact that such methods of communication were not envisioned in 1966 is of little consequence, because the parties must have contemplated when concluding the Covenant the continuation of progress in communication technology (in the same way that television devices represented in 1966 a relatively new form of media, when compared to radio broadcasts). The unique scale, scope and speed of content dissemination over digital platforms may warrant, however, reconsideration through dynamic interpretation of the precise balance that ought to be struck between freedom of expression and the countervailing rights and interests of impacted individuals and societies. Hence, for example, more pro-active content moderation responses than the ones applied to traditional forms of communication might be needed in order to counter the particular risks posed by harmful online contents involving hate speech, defamation, bullying, disinformation and the like.^[125]

Yet, in other contexts, more extensive adjustments to the contents of IHRL treaty norms may be required in order to adapt them to the challenges of digital technology. This might be so because the human interactions, human needs and interests and forms of empowerment and disempowerment caused or facilitated by digital technology were not, and could not have been, envisioned when the relevant IHRL treaty norms were drafted. Moreover, the language of some of these treaty norms is not as ‘future-proof’ as that of article 19, and does not readily lend itself to extension to new technological conditions. Yet, even when the treaty text can technically accommodate new meanings (as is the case with the right to privacy further discussed below), it remains questionable whether such an extension should be legitimately undertaken notwithstanding contrary indications emanating from other semantic sources – lack of intent or expectation that the parties’ substantive obligations under IHRL treaties would undergo a significant expansion to wholly new areas of human activity – and given the risk of creating or further expanding mega-human rights. As I discuss below, the candidate digital human rights to digital self-determination and algorithmic transparency may fall under this category of significant and unanticipated normative change, which will be difficult to read into existing IHRL treaty norms.

Second, the process of treaty making at the global level in the field of IHRL has become extremely difficult to pursue in recent years, due to geopolitical tensions and populist trends, resulting in the process coming to a screeching halt (the last major global human right treaty – the Convention on Enforced Disappearance was concluded in 2006).^[126] Some work on developing new IHRL norms continues at the global level, yet their process of graduation into customary international law or treaty law remains long and uncertain.^[127] Furthermore, while at the regional level, the process of human rights law-making has not stopped,^[128] its impact on the development of new digital human rights law has been limited so far. Most significantly, no protocol to any of the major regional human rights treaties which contains a list of digital human rights has been concluded, or is being seriously considered, it being ‘not(…) advisable at this stage’.^[129] Regional texts in this sub-field typically cover digital rights (as opposed to digital human rights)^[130] or principles relating to the application of human rights in the digital domain (as opposed to specific digital human rights).^[131] This leaves dynamic interpretation as the main vehicle for periodically incorporating digital rights into IHRL treaties.

One relevant paradox that arises in this regard is that the more unlikely it is that new IHRL treaties would be concluded, the greater is the inclination by key stakeholders to maintain that there is no need for such new treaties, since existing treaty language can be construed in a manner that tackles the impacts of digital technology.^[132] This inclination – which might be designed to avoid acknowledging the existence of a normative vacuum^[133] – amounts to a self-fulfilling prophecy: Given the high transaction costs associated with IHRL treaty-making, the claim that new digital human rights treaties are redundant reduces to a minimum the prospects of concluding them. Most normative innovations are then relegated to the realm of dynamic interpretation, notwithstanding the legitimacy concerns and legal challenges associated with such an approach.

Third, a significant part of the effort to mediate between new and emerging digital technology and the regulatory responses required to protect the needs and interests of individuals impacted by digital technology (as well as to harness the new opportunities offered to individuals by digital technology), has been undertaken, up until now, by regional lawmakers through legal instruments that provide individuals with new legal protections in the form of digital rights (such individual beneficiaries are sometimes referred to, in these instruments, as ‘data subjects’). Indeed, digital rights, such as those found in the GDPR,^[134] Council of Europe Convention 108+^[135] or the African Union Malabo Convention,^[136] resemble IHRL norms in their language and in the nature of their engagement with needs and interests that hold great personal significance for individuals, which correspond with basic human rights values, such as dignity, liberty and equality. They are also aimed at generating legal conditions that could better facilitate the application of traditional human rights, such as the right to privacy, in digital contexts. Still, these regional instruments do not confer on their individual beneficiaries full-fledged human rights – that is, they do not purport, by and large, to extend their protection to every individual in the world as the principle of universality of IHRL requires;^[137] nor do they feature a clear intent to ascribe to the enumerated rights the uniquely elevated legal status associated with human rights (eg, inalienability), which catapults them into the corpus of fundamental rights trumping most other legal rights and interests.^[138] Finally, the digital rights introduced by regional instruments are not embedded in the broader regional human rights machinery available to monitor the application of regional human rights treaties, such as the ECHR.

Still, the emergence of this liminal legal status – human rights-like digital rights– is important for the subsequent development of digital human rights. It offers a model for formulating candidate digital human rights, and generates relevant case law – a corpus of legal interpretations – which can be used later in the service of digital human rights. Specifically, adjudication around digital rights offers interpretative points of reference regarding the relationship between digital rights and IHRL that could facilitate reading digital rights into existing IHRL treaties through dynamic interpretation. Indeed, most of the examples I review below feature attempts by regional courts or human rights rapporteurs to explicitly link digital rights to existing IHRL treaty provisions.

3.2 Incorporating Specific Digital Rights

3.3 The Limits of Reading New Digital Rights into Existing Rights

The three examples of incorporation of digital rights into IHRL treaties discussed above illustrate both the potential and the limits of dynamic interpretation exercises undertaken by norm entrepreneurs, such as UN Special Rapporteurs, or CJEU judges or advocates-general. Clearly, there are normative overlaps between digital rights and existing IHRL treaties, and respect for digital rights may obviously facilitate the enjoyment of human rights in an era of increasing dependency of individuals on digital technology. As a result, some aspects of digital rights could be linked – albeit more tenuously in some case than in others – to the language of certain IHRL treaty norms, and perhaps also to the putative intent of the relevant drafters to future-proof certain treaty provisions and to the legitimate expectations of the parties that, in order to remain effective, IHRL treaties should be gradually adapted to changes in technology and society.^[163]

Still, it is also important to acknowledge what reading new digital rights into existing IHRL treaties fails to do, as well as the risks attendant to excessive use of dynamic interpretation − ie, resorting to ultra-dynamic interpretation. Arguably, the three aforementioned digital rights – right to internet access, right to be forgotten and right to a human decision – can be conceptualized as parts of a broader set of claims that only partly overlap with the specific human rights norms with which they were linked in the three cases I discussed above. As a result, their association with specific existing rights might actually circumscribe their evolution into full-fledged digital human rights.

For example, viewing the right to internet access only as a vehicle for exercising freedom of opinion and expression (or a derivative right emanating from freedom of opinion and expression)^[164] ignores other important participatory dimensions of being online, including participating in political life, engaging in social life – including in transnational communities – developing business and educational opportunities, accessing public services etc. Dafna Dror-Shpoliansky and I have remarked in an earlier publication that, for the purposes of enjoying digital rights (and digital human rights), a right to internet access represents the equivalent of the Arendtian ‘right to have rights’.^[165] It is therefore doubtful whether a freedom of opinion and expression framework is capable of capturing these critically important dimensions of internet access. Although La Rue alluded in his 2011 report to the enabling role of freedom of expression for other human rights, and – by implication – to the role of internet access as a gateway to other rights, even this broader framing only covers expressive or informational aspects of internet access. In other words, La Rue created a mediating interpretive prism which might be helpful in providing some degree of human rights protection to the right to internet access, but which could, were it to remain the only available prism, ultimately stifle its long-term development as a key digital human right.

In a similar vein, the right to be forgotten can be associated with a broader digital human right – the right to digital (or informational) self-determination.^[166] This right has been developed in German public law in order to give effect to the legitimate interest of individuals in exercising control and agency with relation to data collected about them.^[167] A number of GDPR rights give effect to specific aspects of this broad digital right: the right to be informed on personal data collection,^[168] the right of access to personal data and the right to be informed about the processing of such data,^[169] the right to rectification,^[170] the right to erasure (right to be forgotten),^[171] the right to restrict or object to data processing^[172] and the right to data portability.^[173] The right to digital self-determination has been recently recognized by the Inter-American Court of Human Rights as an autonomous right that partly overlaps with some existing human rights under the American Convention on Human Rights, including the right to privacy, the right to seek information and the right to judicial remedy.^[174]

While some aspects of the right to digital self-determination indeed overlap with existing IHRL treaty norms, including the protection of reputation implicated by the right to be forgotten,^[175] some other aspects of the right to digital self-determination appear to exceed the scope of existing IHRL treaties. In particular, the notion of individual control over all types of personal data – as opposed to sensitive personal data or information about private life^[176] – seems to be motivated less by traditional privacy concerns about protecting a sphere of intimacy (or prohibiting attacks on reputation),^[177] and more by other concerns about human wellbeing. These include concerns about obstructing self-development through social control exercised over individuals by governments or other public or private actors having unlimited control over their data,^[178] and an understanding of personal data as a projection of one’s self,^[179] enabling individuals to curate their public image.^[180] The right to digital self-determination also suggests the existence of a propriety interest in personal data and reveals concerns about the personal and societal harms associated with commercial data gathering, storing, processing, using and sharing.^[181] Here too, the development of this candidate digital human right exclusively through dynamic interpretation of existing IHRL treaty norms, such as the right to privacy, may restrict the scope of protection the right would eventually afford individuals, and complicate its application.

Finally, the right to a human decision also comprises part of a broader candidate digital human right to algorithmic fairness and transparency,^[182] and perhaps also part of an emerging digital human right to human-to-human interaction.^[183] While automated decision-making raises concerns about discriminatory profiling, arbitrary denial of rights, due process (especially when automation is used in legal proceedings) and excessive collection and misuse of personal data, the main significance of the right might be found elsewhere: In the human dignity implications of having matters affecting important interests of individuals being decided by an algorithm, in the need to facilitate one’s participation in important decisions concerning herself/himself and the demand for fairness and transparency in algorithmic decision-making.^[184] It is quite clear that it would be difficult to address all of these broader objectives through a dynamic interpretation existing IHRL treaties, or through a mega-human right, such as the right to privacy.^[185] Indeed, replacing a human decision maker with an algorithm does not fundamentally change the level of intrusion into personal life per se. Rather, such a replacement mostly relates to considerations of human wellbeing whose relationship to privacy seem to be, at best peripheral in nature.

One possible response to concerns about the constraining effect of dynamic interpretation on the future development of candidate digital human rights is to advocate for an ultra-dynamic interpretation of existing IHRL treaties, with a view to capture within their scope of coverage those broad aspects of the candidate rights which have not yet been linked to existing IHRL treaty norms – eg, internet access as a gateway right, digital self-determination as a broad right to control personal data and the right to algorithmic fairness and transparency. The problem with this approach is that it would likely raise significant legitimacy challenges in connection with all three semantic sources discussed in this article. It would also facilitate the creation or further expansion of mega-human rights, whose development could raise additional practical difficulties.

1. Language – it is difficult to reconcile the text of existing IHRL treaties when read in context, even against an abstract object and purpose to promote human rights, with the protection of new human needs and interests that were clearly not contemplated when the treaties were drafted, and which did not receive, as a result, any expression in their language. This is especially the case when the text does not contain adequate future-proofing language. There is a fundamental difference, in this regard, between applying existing law designed to protect a more or less constant human need or interest, such as to seek, receive and impart information, throughout changes in the technological platforms that enable it (for example, the internet instead of radio or television), and reading into existing IHRL treaties the protection of wholly new needs and interests, which may be morally justified but were never contemplated before, such as exercising power of control over personal data located in the public sphere or advancing transparency in algorithmic decision-making. The farther removed are such new needs and interests from the original needs and interests that the treaty language captured, the more tenuous and less legitimate are attempts to link candidate digital human rights associated with the new needs and interests to existing IHRL treaties.^[186]

Under these circumstances, a resort to dynamic interpretation in order to read into existing IHRL treaties completely new digital rights, broad in their scope of coverage and unique in the needs and interests they protect, would have to be justified on the basis of abstract teleological considerations, such as maintaining the effectiveness of existing treaties by way of incorporating within them new aspects that facilitate the enjoyment of existing human rights (eg, internet access as supporting the right to work),^[187] or by linking the proposed interpretation to general values, which existing rights were also intended to advance (such as human decision as an aspect of human dignity). Still, the more abstract such interpretative considerations are, the greater is the likelihood that courts and other law-interpreters resorting to them would be accused of illegitimately engaging in ultra-dynamic interpretation or in judicial or quasi-judicial legislation.

Furthermore, even if the existing language of an IHRL treaty can accommodate a significant modification of the contents of existing rights, an ultra-dynamic interpretation incorporating therein new digital rights increases the risk of creating new mega-rights, bringing together under one normative umbrella diverse legal norms which purport to advance a multiplicity of normative agendas. As indicated above, the interpretation and application of mega-human rights entail significant practical challenges. What’s more, to the extent that new candidate human rights have normative objects and purposes different from the ones attached to existing human rights, reading the former into the latter might run contrary to the ‘in the light of its object and purpose’ interpretative guidance afforded by the VCLT (provided that the VCLT’s ‘object and purpose’ standard of interpretation applies not only to the treaty a whole, but also to interpretation of specific provisions thereof).^[188]

1. Drafters’ Intentions – In the absence of any specific indication for the drafters’ intention to future-proof existing IHRL treaty provisions, so as to allow courts and other law-interpreters to adapt them by way of interpretation, it is difficult to accept that the drafters intended to protect needs and interests which did not exist at all at the time in which the treaty was concluded.^[189] In some cases, where the normative distance between previously protected needs and interests and new needs and interests is reasonably narrow, such an intention can be presumed as part of the ‘living instrument’ approach: As noted before, the text of article 19 of the ICCPR manifests the drafters’ intention to future-proof freedom of expression protections in light of changes in communication technology, and to adjust accordingly the relationship between the right and restrictions on the right. The same may be said with regard to the scope of coverage of the right to privacy and practices of constant and ubiquitous data collection (involving the mass collection of personal data, including sensitive personal information, which could subject individuals to the uneasy sense of living in a panopticon).^[190] Such technology might not been envisioned at the time in which the ICCPR was drafted, but it can be presumed to fall within the drafters’ intent to effectively protect private life across time, including from potential threats posed by new and emerging digital technologies that erode existing privacy safeguards.^[191] Still, completely new needs and interests, such as regulating human interactions with AI systems, do not appear to plausibly come within the purview of any need or interest envisioned by the drafters of IHRL treaties.

2. Parties’ legitimate expectations – One way to address the limits of existing IHRL treaty language and the inability to rely on the drafters’ intentions regarding the scope of protection afforded by the treaty is to factor into the legal analysis the legitimate expectations of the parties to the relevant instruments, as a way to either constrain or expand the scope of the treaty. Such legitimate expectations would need to appertain both to the substantive aspects of incorporating new digital rights into existing IHRL treaties and to the role of courts and other law-interpreters in advancing such an interpretation. As for the first aspect, there is clearly broad support across states for adapting existing human rights to digital contexts, including, for example, support for the proposition that freedom of expression is applicable online,^[192] and that the right to privacy also covers digital privacy (eg, it limits practices of online surveillance and allows for control by individuals over sensitive personal data).^[193] There may also be support by certain stakeholders for linking to existing IHRL treaty provisions, such as the right to privacy, certain digital rights, like the right to be forgotten^[194] and the right to a human decision.^[195] It is interesting to note, however, that a number of civil society groups questioned in their intervention in the ECtHR Hurbain case whether the right to be forgotten is part and parcel of international law,^[196] and that the CJEU appears to have been reluctant to link the right to a human decision to the EU Charter of Fundamental Rights. This perhaps suggests that consensus on the relationship between digital rights and existing human rights has not yet been fully attained.

It is less clear, however, whether the legitimate expectations of the parties allow for incorporating into exiting IHRL treaty provisions, by way of ultra-dynamic interpretation, the overarching candidate digital human rights which the three digital rights discussed above comprise a part of (Internet access as a gateway right, right to digital self-determination and right to algorithmic fairness and transparency) or digital rights entailing onerous positive obligations (such as developing cutting-edge digital infrastructure as part of the right to access the Internet). It is also questionable whether state parties would accept that law-interpreters should be given a carte blanche to significantly expand existing rights by way of resorting to ultra-dynamic interpretation.^[197]