Vue normale

Il y a de nouveaux articles disponibles, cliquez pour rafraîchir la page.
À partir d’avant-hierFlux principal

UN Cybercrime Draft Convention Dangerously Expands State Surveillance Powers Without Robust Privacy, Data Protection Safeguards

This is the third post in a series highlighting flaws in the proposed UN Cybercrime Convention. Check out Part I, our detailed analysis on the criminalization of security research activities, and Part II, an analysis of the human rights safeguards.

As we near the final negotiating session for the proposed UN Cybercrime Treaty, countries are running out of time to make much-needed improvements to the text. From July 29 to August 9, delegates in New York aim to finalize a convention that could drastically reshape global surveillance laws. The current draft favors extensive surveillance, establishes weak privacy safeguards, and defers most protections against surveillance to national laws—creating a dangerous avenue that could be exploited by countries with varying levels of human rights protections.

The risk is clear: without robust privacy and human rights safeguards in the actual treaty text, we will see increased government overreach, unchecked surveillance, and unauthorized access to sensitive data—leaving individuals vulnerable to violations, abuses, and transnational repression. And not just in one country.  Weaker safeguards in some nations can lead to widespread abuses and privacy erosion because countries are obligated to share the “fruits” of surveillance with each other. This will worsen disparities in human rights protections and create a race to the bottom, turning global cooperation into a tool for authoritarian regimes to investigate crimes that aren’t even crimes in the first place.

Countries that believe in the rule of law must stand up and either defeat the convention or dramatically limit its scope, adhering to non-negotiable red lines as outlined by over 100 NGOs. In an uncommon alliance, civil society and industry agreed earlier this year in a joint letter urging governments to withhold support for the treaty in its current form due to its critical flaws.

Background and Current Status of the UN Cybercrime Convention Negotiations

The UN Ad Hoc Committee overseeing the talks and preparation of a final text is expected to consider a revised but still-flawed text in its entirety, along with the interpretative notes, during the first week of the session, with a focus on all provisions not yet agreed ad referendum.[1] However, in keeping with the principle in multilateral negotiations that “nothing is agreed until everything is agreed,” any provisions of the draft that have already been agreed could potentially be reopened. 

The current text reveals significant disagreements among countries on crucial issues like the convention's scope and human rights protection. Of course the text could also get worse. Just when we thought Member States had removed many concerning crimes, they could reappear. The Ad-Hoc Committee Chair’s General Assembly resolution includes two additional sessions to negotiate not more protections, but the inclusion of more crimes. The resolution calls for “a draft protocol supplementary to the Convention, addressing, inter alia, additional criminal offenses.” Nevertheless, some countries still expect the latest draft to be adopted.

In this third post, we highlight the dangers of the currently proposed UN Cybercrime Convention's broad definition of "electronic data" and inadequate privacy and data protection safeguards.Together, these create the conditions for severe human rights abuses, transnational repression, and inconsistencies across countries in human rights protections.

A Closer Look to the Definition of Electronic Data

The proposed UN Cybercrime Convention significantly expands state surveillance powers under the guise of combating cybercrime. Chapter IV grants extensive government authority to monitor and access digital systems and data, categorizing data into communications data: subscriber data, traffic data, and content data. But it also makes use of a catch-all category called "electronic data." Article 2(b) defines electronic data as "any representation of facts, information, or concepts in a form suitable for processing in an information and communications technology system, including a program suitable to cause an information and communications technology system to perform a function."

"Electronic data," is eligible for three surveillance powers: preservation orders (Article 25), production orders (Article 27), and search and seizure (Article 28). Unlike the other traditional categories of traffic data, subscriber data and content data, "electronic data" refers to any data stored, processed, or transmitted electronically, regardless of whether it has been communicated to anyone. This includes documents saved on personal computers or notes stored on digital devices. In essence, this means that private unshared thoughts and information are no longer safe. Authorities can compel the preservation, production, or seizure of any electronic data, potentially turning personal devices into spy vectors regardless of whether the information has been communicated.

This is delicate territory, and it deserves careful thought and real protection—many of us now use our devices to keep our most intimate thoughts and ideas, and many of us also use tools like health and fitness tools in ways that we do not intend to share. This includes data stored on devices, such as face scans and smart home device data, if they remain within the device and are not transmitted. Another example could be photos that someone takes on a device but doesn't share with anyone. This category threatens to turn our most private thoughts and actions over to spying governments, both our own and others. 

And the problem is worse when we consider emerging technologies. The sensors in smart devices, AI, and augmented reality glasses, can collect a wide array of highly sensitive data. These sensors can record involuntary physiological reactions to stimuli, including eye movements, facial expressions, and heart rate variations. For example, eye-tracking technology can reveal what captures a user's attention and for how long, which can be used to infer interests, intentions, and even emotional states. Similarly, voice analysis can provide insights into a person's mood based on tone and pitch, while body-worn sensors might detect subtle physical responses that users themselves are unaware of, such as changes in heart rate or perspiration levels.

These types of data are not typically communicated through traditional communication channels like emails or phone calls (which would be categorized as content or traffic data). Instead, they are collected, stored, and processed locally on the device or within the system, fitting the broad definition of "electronic data" as outlined in the draft convention.

Such data likely has been harder to obtain because it may have not been communicated to or possessed by any communications intermediary or system. So it’s an  example of how the broad term "electronic data" increases the kinds (and sensitivity) of information about us that can be targeted by law enforcement through production orders or by search and seizure powers. These emerging technology uses are their own category, but they are most like "content" in communications surveillance, which usually has high protection. “Electronic data” must have equal protection as “content” of communication, and be subject to ironclad data protection safeguards, which the propose treaty fails to provide, as we will explain below.

The Specific Safeguard Problems

Like other powers in the draft convention, the broad powers related to "electronic data" don't come with specific limits to protect fair trial rights. 

Missing Safeguards

For example, many countries' have various kinds of information that is protected by a legal “privilege” against surveillance: attorney-client privilege, the spousal privilege, the priest-penitent privilege, doctor-patient privileges, and many kinds of protections for confidential business information and trade secrets. Many countries, also give additional protections for journalists and their sources. These categories, and more, provide varying degrees of extra requirements before law enforcement may access them using production orders or search-and-seizure powers, as well as various protections after the fact, such as preventing their use in prosecutions or civil actions. 

Similarly, the convention lacks clear safeguards to prevent authorities from compelling individuals to provide evidence against themselves. These omissions raise significant red flags about the potential for abuse and the erosion of fundamental rights when a treaty text involves so many countries with a high disparity of human rights protections.

The lack of specific protections for criminal defense is especially troubling. In many legal systems, defense teams have certain protections to ensure they can effectively represent their clients, including access to exculpatory evidence and the protection of defense strategies from surveillance. However, the draft convention does not explicitly protect these rights, which both misses the chance to require all countries to provide these minimal protections and potentially further undermines the fairness of criminal proceedings and the ability of suspects to mount an effective defense in countries that either don’t provide those protections or where they are not solid and clear.

Even the State “Safeguards” in Article 24 are Grossly Insufficient

Even where the convention’s text discusses “safeguards,” the convention doesn’t actually protect people. The “safeguard” section, Article 24, fails in several obvious ways: 

Dependence on Domestic Law: Article 24(1) makes safeguards contingent on domestic law, which can vary significantly between countries. This can result in inadequate protections in states where domestic laws do not meet high human rights standards. By deferring safeguards to national law, Article 24 weakens these protections, as national laws may not always provide the necessary safeguards. It also means that the treaty doesn’t raise the bar against invasive surveillance, but rather confirms even the lowest protections.

A safeguard that bends to domestic law isn't a safeguard at all if it leaves the door open for abuses and inconsistencies, undermining the protection it's supposed to offer.

Discretionary Safeguards: Article 24(2) uses vague terms like “as appropriate,” allowing states to interpret and apply safeguards selectively. This means that while the surveillance powers in the convention are mandatory, the safeguards are left to each state’s discretion. Countries decide what is “appropriate” for each surveillance power, leading to inconsistent protections and potential weakening of overall safeguards.

Lack of Mandatory Requirements: Essential protections such as prior judicial authorization, transparency, user notification, and the principle of legality, necessity and non-discrimination are not explicitly mandated. Without these mandatory requirements, there is a higher risk of misuse and abuse of surveillance powers.

No Specific Data Protection Principles: As we noted above, the proposed treaty does not include specific safeguards for highly sensitive data, such as biometric or privileged data. This oversight leaves such information vulnerable to misuse.

Inconsistent Application: The discretionary nature of the safeguards can lead to their inconsistent application, exposing vulnerable populations to potential rights violations. Countries might decide that certain safeguards are unnecessary for specific surveillance methods, which the treaty allows, increasing the risk of abuse.

Finally, Article 23(4) of Chapter IV authorizes the application of Article 24 safeguards to specific powers within the international cooperation chapter (Chapter V). However, significant powers in Chapter V, such as those related to law enforcement cooperation (Article 47) and the 24/7 network (Article 41) do not specifically cite the corresponding Chapter IV powers and so may not be covered by Article 24 safeguards.

Search and Seizure of Stored Electronic Data

The proposed UN Cybercrime Convention significantly expands government surveillance powers, particularly through Article 28, which deals with the search and seizure of electronic data. This provision grants authorities sweeping abilities to search and seize data stored on any computer system, including personal devices, without clear, mandatory privacy and data protection safeguards. This poses a serious threat to privacy and data protection.

Article 28(1) allows authorities to search and seize any “electronic data” in an information and communications technology (ICT) system or data storage medium. It lacks specific restrictions, leaving much to the discretion of national laws. This could lead to significant privacy violations as authorities might access all files and data on a suspect’s personal computer, mobile device, or cloud storage account—all without clear limits on what may be targeted or under what conditions.

Article 28(2) permits authorities to search additional systems if they believe the sought data is accessible from the initially searched system. While judicial authorization should be a requirement to assess the necessity and proportionality of such searches, Article 24 only mandates “appropriate conditions and safeguards” without explicit judicial authorization. In contrast, U.S. law under the Fourth Amendment requires search warrants to specify the place to be searched and the items to be seized—preventing unreasonable searches and seizures.

Article 28(3) empowers authorities to seize or secure electronic data, including making and retaining copies, maintaining its integrity, and rendering it inaccessible or removing it from the system. For publicly accessible data, this takedown process could infringe on free expression rights and should be explicitly subject to free expression standards to prevent abuse.

Article 28(4) requires countries to have laws that allow authorities to compel anyone who knows how a particular computer or device works to provide necessary information to access it. This could include asking a tech expert or an engineer to help unlock a device or explain its security features. This is concerning because it might force people to help law enforcement in ways that could compromise security or reveal confidential information. For example, an engineer could be required to disclose a security flaw that hasn't been fixed, or to provide encryption keys that protect data, which could then be misused. The way it is written, it could be interpreted to include disproportionate orders that can lead to forcing persons to disclose a vulnerability to the government that hasn’t been fixed. It could also imply forcing people to disclose encryption keys such as signing keys on the basis that these are “the necessary information to enable” some form of surveillance.

Privacy International and EFF strongly recommend Article 28.4 be removed in its entirety. Instead, it has been agreed ad referendum. At least, the drafters must include material in the explanatory memorandum that accompanies the draft Convention to clarify limits to avoid forcing technologists to reveal confidential information or do work on behalf of law enforcement against their will. Once again, it would also be appropriate to have clear legal standards about how law enforcement can be authorized to seize and look through people’s private devices.

In general, production and search and seizure orders might be used to target tech companies' secrets, and require uncompensated labor by technologists and tech companies, not because they are evidence of crime but because they can be used to enhance law enforcement's technical capabilities.

Domestic Expedited Preservation Orders of Electronic Data

Article 25 on preservation orders, already agreed ad referendum, is especially problematic. It’s very broad, and will result in individuals’ data being preserved and available for use in prosecutions far more than needed. It also fails to include necessary safeguards to avoid abuse of power. By allowing law enforcement to demand preservation with no factual justification, it risks spreading familiar deficiencies in U.S. law worldwide.

Article 25 requires each country to create laws or other measures that let authorities quickly preserve specific electronic data, particularly when there are grounds to believe that such data is at risk of being lost or altered.

Article 25(2) ensures that when preservation orders are issued, the person or entity in possession of the data must keep it for up to 90 days, giving authorities enough time to obtain the data through legal channels, while allowing this period to be renewed. There is no specified limit on the number of times the order can be renewed, so it can potentially be reimposed indefinitely.

Preservation orders should be issued only when they’re absolutely necessary, but Article 24 does not mention the principle of necessity and lacks individual notice and explicit grounds requirements and statistical transparency obligations.

The article must limit the number of times preservation orders may be renewed to prevent indefinite data preservation requirements. Each preservation order renewal must require a demonstration of continued necessity and factual grounds justifying continued preservation.

Article 25(3) also compels states to adopt laws that enable gag orders to accompany preservation orders, prohibiting service providers or individuals from informing users that their data was subject to such an order. The duration of such a gag order is left up to domestic legislation.

As with all other gag orders, the confidentiality obligation should be subject to time limits and only be available to the extent that disclosure would demonstrably threaten an investigation or other vital interest. Further, individuals whose data was preserved should be notified when it is safe to do so without jeopardizing an investigation. Independent oversight bodies must oversee the application of preservation orders.

Indeed, academics such as prominent law professor and former U.S. Department of Justice lawyer Orin S. Kerr have criticized similar U.S. data preservation practices under 18 U.S.C. § 2703(f) for allowing law enforcement agencies to compel internet service providers to retain all contents of an individual's online account without their knowledge, any preliminary suspicion, or judicial oversight. This approach, intended as a temporary measure to secure data until further legal authorization is obtained, lacks the foundational legal scrutiny typically required for searches and seizures under the Fourth Amendment, such as probable cause or reasonable suspicion.

The lack of explicit mandatory safeguards raise similar concerns about Article 25 of the proposed UN convention. Kerr argues that these U.S. practices constitute a "seizure" under the Fourth Amendment, indicating that such actions should be justified by probable cause or, at the very least, reasonable suspicion—criteria conspicuously absent in the current draft of the UN convention.

By drawing on Kerr's analysis, we see a clear warning: without robust safeguards— including an explicit grounds requirement, prior judicial authorization, explicit notification to users, and transparency—preservation orders of electronic data proposed under the draft UN Cybercrime Convention risk replicating the problematic practices of the U.S. on a global scale.

Production Orders of Electronic Data

Article 27(a)’s treatment of “electronic data” in production orders, in light of the draft convention’s broad definition of the term, is especially problematic.

This article, which has already been agreed ad referendum, allows production orders to be issued to custodians of electronic data, requiring them to turn over copies of that data. While demanding customer records from a company is a traditional governmental power, this power is dramatically increased in the draft convention.

As we explain above, the extremely broad definition of electronic data, which is often sensitive in nature, raises new and significant privacy and data protection concerns, as it permits authorities to access potentially sensitive information without immediate oversight and prior judicial authorization. The convention needs instead to require prior judicial authorization before such information can be demanded from the companies that hold it. 

This ensures that an impartial authority assesses the necessity and proportionality of the data request before it is executed. Without mandatory data protection safeguards for the processing of personal data, law enforcement agencies might collect and use personal data without adequate restrictions, thereby risking the exposure and misuse of personal information.

The text of the convention fails to include these essential data protection safeguards. To protect human rights, data should be processed lawfully, fairly, and in a transparent manner in relation to the data subject. Data should be collected for specified, explicit, and legitimate purposes and not further processed in a manner that is incompatible with those purposes. 

Data collected should be adequate, relevant, and limited to what is necessary to the purposes for which they are processed. Authorities should request only the data that is essential for the investigation. Production orders should clearly state the purpose for which the data is being requested. Data should be kept in a format that permits identification of data subjects for no longer than is necessary for the purposes for which the data is processed. None of these principles are present in Article 27(a) and they must be. 

International Cooperation and Electronic Data

The draft UN Cybercrime Convention includes significant provisions for international cooperation, extending the reach of domestic surveillance powers across borders, by one state on behalf of another state. Such powers, if not properly safeguarded, pose substantial risks to privacy and data protection. 

  • Article 42 (1) (“International cooperation for the purpose of expedited preservation of stored electronic data”) allows one state to ask another to obtain preservation of “electronic data” under the domestic power outlined in Article 25. 
  • Article 44 (1) (“Mutual legal assistance in accessing stored electronic data”) allows one state to ask another “to search or similarly access, seize or similarly secure, and disclose electronic data,” presumably using powers similar to those under Article 28, although that article is not referenced in Article 44. This specific provision, which has not yet been agreed ad referendum, enables comprehensive international cooperation in accessing stored electronic data. For instance, if Country A needs to access emails stored in Country B for an ongoing investigation, it can request Country B to search and provide the necessary data.

Countries Must Protect Human Rights or Reject the Draft Treaty

The current draft of the UN Cybercrime Convention is fundamentally flawed. It dangerously expands surveillance powers without robust checks and balances, undermines human rights, and poses significant risks to marginalized communities. The broad and vague definitions of "electronic data," coupled with weak privacy and data protection safeguards, exacerbate these concerns.

Traditional domestic surveillance powers are particularly concerning as they underpin international surveillance cooperation. This means that one country can easily comply with the requests of another, which if not adequately safeguarded, can lead to widespread government overreach and human rights abuses. 

Without stringent data protection principles and robust privacy safeguards, these powers can be misused, threatening human rights defenders, immigrants, refugees, and journalists. We urgently call on all countries committed to the rule of law, social justice, and human rights to unite against this dangerous draft. Whether large or small, developed or developing, every nation has a stake in ensuring that privacy and data protection are not sacrificed. 

Significant amendments must be made to ensure these surveillance powers are exercised responsibly and protect privacy and data protection rights. If these essential changes are not made, countries must reject the proposed convention to prevent it from becoming a tool for human rights violations or transnational repression.

[1] In the context of treaty negotiations, "ad referendum" means that an agreement has been reached by the negotiators, but it is subject to the final approval or ratification by their respective authorities or governments. It signifies that the negotiators have agreed on the text, but the agreement is not yet legally binding until it has been formally accepted by all parties involved.

The UN Cybercrime Draft Convention is a Blank Check for Surveillance Abuses

This is the second post in a series highlighting the problems and flaws in the proposed UN Cybercrime Convention. Check out our detailed analysis on the criminalization of security research activities under the proposed convention.

The United Nations Ad Hoc Committee is just weeks away from finalizing a too-broad Cybercrime Draft Convention. This draft would normalize unchecked domestic surveillance and rampant government overreach, allowing serious human rights abuses around the world.

The latest draft of the convention—originally spearheaded by Russia but since then the subject of two and a half years of negotiations—still authorizes broad surveillance powers without robust safeguards and fails to spell out data protection principles essential to prevent government abuse of power.

As the August 9 finalization date approaches, Member States have a last chance to address the convention’s lack of safeguards: prior judicial authorization, transparency, user notification, independent oversight, and data protection principles such as transparency, minimization, notification to users, and purpose limitation. If left as is, it can and will be wielded as a tool for systemic rights violations.

Countries committed to human rights and the rule of law must unite to demand stronger data protection and human rights safeguards or reject the treaty altogether. These domestic surveillance powers are critical as they underpin international surveillance cooperation.

EFF’s Advocacy for Human Rights Safeguards

EFF has consistently advocated for human rights safeguards to be a baseline for both the criminal procedural measures and international cooperation chapters. The collection and use of digital evidence can implicate human rights, including privacy, free expression, fair trial, and data protection. Strong safeguards are essential to prevent government abuse.

Regrettably, many states already fall short in these regards. In some cases, surveillance laws have been used to justify overly broad practices that disproportionately target individuals or groups based on their political views—particularly ethnic and religious groups. This leads to the suppression of free expression and association, the silencing of dissenting voices, and discriminatory practices. Examples of these abuses include covert surveillance of internet activity without a warrant, using technology to track individuals in public, and monitoring private communications without legal authorization, oversight, or safeguards.

The Special Rapporteur on the rights to freedom of peaceful assembly and of association has already sounded the alarm about the dangers of current surveillance laws, urging states to revise and amend these laws to comply with international human rights norms and standards governing the rights to privacy, free expression, peaceful assembly, and freedom of association. The UN Cybercrime Convention must be radically amended to avoid entrenching and expanding these existing abuses globally. If not amended, it must be rejected outright.

How the Convention Fails to Protect Human Rights in Domestic Surveillance

The idea that checks and balances are essential to avoid abuse of power is a basic “Government 101” concept. Yet throughout the negotiation process, Russia and its allies have sought to chip away at the already-weakened human rights safeguards and conditions outlined in Article 24 of the proposed Convention. 

Article 24 as currently drafted requires that every country that agrees to this convention must ensure that when it creates, uses, or applies the surveillance powers and procedures described in the domestic procedural measures, it does so under its own laws. These laws must protect human rights and comply with international human rights law. The principle of proportionality must be respected, meaning any surveillance measures should be appropriate and not excessive in relation to the legitimate aim pursued.

Why Article 24 Falls Short?

1. The Critical Missing Principles

While incorporation of the principle of proportionality in Article 24(1) is commendable, the article still fails to explicitly mention the principles of legality, necessity, and non-discrimination, which hold equivalent status to proportionality in human rights law relative to surveillance activities. A primer:

  • The principle of legality requires that restrictions on human rights including the right to privacy be authorized by laws that are clear, publicized, precise, and predictable, ensuring individuals understand what conduct might lead to restrictions on their human rights.
  • The principles of necessity and proportionality ensure that any interference with human rights is demonstrably necessary to achieving a legitimate aim and only include measures that are proportionate to that aim.
  • The principle of non-discrimination requires that laws, policies and human rights obligations be applied equally and fairly to all individuals, without any form of discrimination based on race, color, sex, language, religion, political or other opinion, national or social origin, property, birth, or other status, including the application of surveillance measures.

Without including all these principles, the safeguards are incomplete and inadequate, increasing the risk of misuse and abuse of surveillance powers.

2. Inadequate Specific Safeguards 

Article 24(2) requires countries to include, where “appropriate,” specific safeguards like:

  • judicial or independent review, meaning surveillance actions must be reviewed or authorized by a judge or an independent regulator.
  • the right to an effective remedy, meaning people must have ways to challenge or seek remedy if their rights are violated.
  • justification and limits, meaning there must be clear reasons for using surveillance and limits on how much surveillance can be done and for how long.

Article 24 (2) introduces three problems:

2.1 The Pitfalls of Making Safeguards Dependent on Domestic Law

Although these safeguards are mentioned, making them contingent on domestic law can vastly weaken their effectiveness, as national laws vary significantly and many of them won’t provide adequate protections. 

2.2 The Risk of Ambiguous Terms Allowing Cherry-Picked Safeguards

The use of vague terms like “as appropriate” in describing how safeguards will apply to individual procedural powers allows for varying interpretations, potentially leading to weaker protections for certain types of data in practice. For example, many states provide minimal or no safeguards for accessing subscriber data or traffic data despite the intrusiveness of resulting surveillance practices. These powers have been used to identify anonymous online activity, to locate and track people, and to map people’s contacts. By granting states broad discretion to decide which safeguards to apply to different surveillance powers, the convention fails to ensure the text will be implemented in accordance with human rights law. Without clear mandatory requirements, there is a real risk that essential protections will be inadequately applied or omitted altogether for certain specific powers, leaving vulnerable populations exposed to severe rights violations. Essentially, a country could just decide that some human rights safeguards are superfluous for a particular kind or method of surveillance, and dispense with them, opening the door for serious human rights abuses.

2.3 Critical Safeguards Missing from Article 24(2)

The need for prior judicial authorization, for transparency, and for user notification is critical to any effective and proportionate surveillance power, but not included in Article 24(2).

Prior judicial authorization means that before any surveillance action is taken, it must be approved by a judge. This ensures an independent assessment of the necessity and proportionality of the surveillance measure before it is implemented. Although Article 24 mentions judicial or other independent review, it lacks a requirement for prior judicial authorization. This is a significant omission that increases the risk of abuse and infringement on individuals' rights. Judicial authorization acts as a critical check on the powers of law enforcement and intelligence agencies.

Transparency involves making the existence and extent of surveillance measures known to the public; people must be fully informed of the laws and practices governing surveillance so that they can hold authorities accountable. Article 24 lacks explicit provisions for transparency, so surveillance measures could be conducted in secrecy, undermining public trust and preventing meaningful oversight. Transparency is essential for ensuring that surveillance powers are not misused and that individuals are aware of how their data might be collected and used.

User notification means that individuals who are subjected to surveillance are informed about it, either at the time of the surveillance or afterward when it no longer jeopardizes the investigation. The absence of a user notification requirement in Article 24(2) deprives people of the opportunity to challenge the legality of the surveillance or seek remedies for any violations of their rights. User notification is a key component of protecting individuals’ rights to privacy and due process. It may be delayed, with appropriate justification, but it must still eventually occur and the convention must recognize this.

Independent oversight involves monitoring by an independent body to ensure that surveillance measures comply with the law and respect human rights. This body can investigate abuses, provide accountability, and recommend corrective actions. While Article 24 mentions judicial or independent review, it does not establish a clear mechanism for ongoing independent oversight. Effective oversight requires a dedicated, impartial body with the authority to review surveillance activities continuously, investigate complaints, and enforce compliance. The lack of a robust oversight mechanism weakens the framework for protecting human rights and allows potential abuses to go unchecked.

Conclusion

While it’s somewhat reassuring that Article 24 acknowledges the binding nature of human rights law and its application to surveillance powers, it is utterly unacceptable how vague the article remains about what that actually means in practice. The “as appropriate” clause is a dangerous loophole, letting states implement intrusive powers with minimal limitations and no prior judicial authorization, only to then disingenuously claim this was “appropriate.” This is a blatant invitation for abuse. There’s nothing “appropriate” about this, and the convention must be unequivocally clear about that.

This draft in its current form is an egregious betrayal of human rights and an open door to unchecked surveillance and systemic abuses. Unless these issues are rectified, Member States must recognize the severe flaws and reject this dangerous convention outright. The risks are too great, the protections too weak, and the potential for abuse too high. It’s long past time to stand firm and demand nothing less than a convention that genuinely safeguards human rights.

Check out our detailed analysis on the criminalization of security research activities under the UN Cybercrime Convention. Stay tuned for our next post, where we'll explore other critical areas affected by the convention, including its scope and human rights safeguards.




In Historic Victory for Human Rights in Colombia, Inter-American Court Finds State Agencies Violated Human Rights of Lawyers Defending Activists

In a landmark ruling for fundamental freedoms in Colombia, the Inter-American Court of Human Rights found that for over two decades the state government harassed, surveilled, and persecuted members of a lawyer’s group that defends human rights defenders, activists, and indigenous people, putting the attorneys’ lives at risk. 

The ruling is a major victory for civil rights in Colombia, which has a long history of abuse and violence against human rights defenders, including murders and death threats. The case involved the unlawful and arbitrary surveillance of members of the Jose Alvear Restrepo Lawyers Collective (CAJAR), a Colombian human rights organization defending victims of political persecution and community activists for over 40 years.

The court found that since at least 1999, Colombian authorities carried out a constant campaign of pervasive secret surveillance of CAJAR members and their families. That state violated their rights to life, personal integrity, private life, freedom of expression and association, and more, the Court said. It noted the particular impact experienced by women defenders and those who had to leave the country amid threat, attacks, and harassment for representing victims.  

The decision is the first by the Inter-American Court to find a State responsible for violating the right to defend human rights. The court is a human rights tribunal that interprets and applies the American Convention on Human Rights, an international treaty ratified by over 20 states in Latin America and the Caribbean. 

In 2022, EFF, Article 19, Fundación Karisma, and Privacy International, represented by Berkeley Law’s International Human Rights Law Clinic, filed an amicus brief in the case. EFF and partners urged the court to rule that Colombia’s legal framework regulating intelligence activity and the surveillance of CAJAR and their families violated a constellation of human rights and forced them to limit their activities, change homes, and go into exile to avoid violence, threats, and harassment. 

Colombia's intelligence network was behind abusive surveillance practices in violation of the American Convention and did not prevent authorities from unlawfully surveilling, harassing, and attacking CAJAR members, EFF told the court. Even after Colombia enacted a new intelligence law, authorities continued to carry out unlawful communications surveillance against CAJAR members, using an expansive and invasive spying system to target and disrupt the work of not just CAJAR but other human rights defenders and journalists

In examining Colombia’s intelligence law and surveillance actions, the court elaborated on key Inter-American and other international human rights standards, and advanced significant conclusions for the protection of privacy, freedom of expression, and the right to defend human rights. 

The court delved into criteria for intelligence gathering powers, limitations, and controls. It highlighted the need for independent oversight of intelligence activities and effective remedies against arbitrary actions. It also elaborated on standards for the collection, management, and access to personal data held by intelligence agencies, and recognized the protection of informational self-determination by the American Convention. We highlight some of the most important conclusions below.

Prior Judicial Order for Communications Surveillance and Access to Data

The court noted that actions such as covert surveillance, interception of communications, or collection of personal data constitute undeniable interference with the exercise of human rights, requiring precise regulations and effective controls to prevent abuse from state authorities. Its ruling recalled European Court of Human Rights’ case law establishing thatthe mere existence of legislation allowing for a system of secret monitoring […] constitutes a threat to 'freedom of communication among users of telecommunications services and thus amounts in itself to an interference with the exercise of rights'.” 

Building on its ruling in the case Escher et al. vs Brazil, the Inter-American Court stated that

“[t]he effective protection of the rights to privacy and freedom of thought and expression, combined with the extreme risk of arbitrariness posed by the use of surveillance techniques […] of communications, especially in light of existing new technologies, leads this Court to conclude that any measure in this regard (including interception, surveillance, and monitoring of all types of communication […]) requires a judicial authority to decide on its merits, while also defining its limits, including the manner, duration, and scope of the authorized measure.” (emphasis added) 

According to the court, judicial authorization is needed when intelligence agencies intend to request personal information from private companies that, for various legitimate reasons, administer or manage this data. Similarly, prior judicial order is required for “surveillance and tracking techniques concerning specific individuals that entail access to non-public databases and information systems that store and process personal data, the tracking of users on the computer network, or the location of electronic devices.”  

The court said that “techniques or methods involving access to sensitive telematic metadata and data, such as email and metadata of OTT applications, location data, IP address, cell tower station, cloud data, GPS and Wi-Fi, also require prior judicial authorization.” Unfortunately, the court missed the opportunity to clearly differentiate between targeted and mass surveillance to explicitly condemn the latter.

The court had already recognized in Escher that the American Convention protects not only the content of communications but also any related information like the origin, duration, and time of the communication. But legislation across the region provides less protection for metadata compared to content. We hope the court's new ruling helps to repeal measures allowing state authorities to access metadata without a previous judicial order.

Indeed, the court emphasized that the need for a prior judicial authorization "is consistent with the role of guarantors of human rights that corresponds to judges in a democratic system, whose necessary independence enables the exercise of objective control, in accordance with the law, over the actions of other organs of public power.” 

To this end, the judicial authority is responsible for evaluating the circumstances around the case and conducting a proportionality assessment. The judicial decision must be well-founded and weigh all constitutional, legal, and conventional requirements to justify granting or denying a surveillance measure. 

Informational Self-Determination Recognized as an Autonomous Human Right 

In a landmark outcome, the court asserted that individuals are entitled to decide when and to what extent aspects of their private life can be revealed, which involves defining what type of information, including their personal data, others may get to know. This relates to the right of informational self-determination, which the court recognized as an autonomous right protected by the American Convention. 

“In the view of the Inter-American Court, the foregoing elements give shape to an autonomous human right: the right to informational self-determination, recognized in various legal systems of the region, and which finds protection in the protective content of the American Convention, particularly stemming from the rights set forth in Articles 11 and 13, and, in the dimension of its judicial protection, in the right ensured by Article 25.”  

The protections that Article 11 grant to human dignity and private life safeguard a person's autonomy and the free development of their personality. Building on this provision, the court affirmed individuals’ self-determination regarding their personal information. In combination with the right to access information enshrined in Article 13, the court determined that people have the right to access and control their personal data held in databases. 

The court has explained that the scope of this right includes several components. First, people have the right to know what data about them are contained in state records, where the data came from, how it got there, the purpose for keeping it, how long it’s been kept, whether and why it’s being shared with outside parties, and how it’s being processed. Next is the right to rectify, modify, or update their data if it is inaccurate, incomplete, or outdated. Third is the right to delete, cancel, and suppress their data in justified circumstances. Fourth is the right to oppose the processing of their data also in justified circumstances, and fifth is the right to data portability as regulated by law. 

According to the court, any exceptions to the right of informational self-determination must be legally established, necessary, and proportionate for intelligence agencies to carry out their mandate. In elaborating on the circumstances for full or partial withholding of records held by intelligence authorities, the court said any restrictions must be compatible with the American Convention. Holding back requested information is always exceptional, limited in time, and justified according to specific and strict cases set by law. The protection of national security cannot serve as a blanket justification for denying access to personal information. “It is not compatible with Inter-American standards to establish that a document is classified simply because it belongs to an intelligence agency and not on the basis of its content,” the court said.  

The court concluded that Colombia violated CAJAR members’ right to informational self -determination by arbitrarily restricting their ability to access and control their personal data within public bodies’ intelligence files.

The Vital Protection of the Right to Defend Human Rights

The court emphasized the autonomous nature of the right to defend human rights, finding that States must ensure people can freely, without limitations or risks of any kind, engage in activities aimed at the promotion, monitoring, dissemination, teaching, defense, advocacy, or protection of universally recognized human rights and fundamental freedoms. The ruling recognized that Colombia violated the CAJAR members' right to defend human rights.

For over a decade, human rights bodies and organizations have raised alarms and documented the deep challenges and perils that human rights defenders constantly face in the Americas. In this ruling, the court importantly reiterated their fundamental role in strengthening democracy. It emphasized that this role justifies a special duty of protection by States, which must establish adequate guarantees and facilitate the necessary means for defenders to freely exercise their activities. 

Therefore, proper respect for human rights requires States’ special attention to actions that limit or obstruct the work of defenders. The court has emphasized that threats and attacks against human rights defenders, as well as the impunity of perpetrators, have not only an individual but also a collective effect, insofar as society is prevented from knowing the truth about human rights violations under the authority of a specific State. 

Colombia’s Intelligence Legal Framework Enabled Arbitrary Surveillance Practices 

In our amicus brief, we argued that Colombian intelligence agents carried out unlawful communications surveillance of CAJAR members under a legal framework that failed to meet international human rights standards. As EFF and allies elaborated a decade ago on the Necessary and Proportionate principles, international human rights law provides an essential framework for ensuring robust safeguards in the context of State communications surveillance, including intelligence activities. 

In the brief, we bolstered criticism made by CAJAR, Centro por la Justicia y el Derecho Internacional (CEJIL), and the Inter-American Commission on Human Rights, challenging Colombia’s claim that the Intelligence Law enacted in 2013 (Law n. 1621) is clear and precise, fulfills the principles of legality, proportionality, and necessity, and provides sufficient safeguards. EFF and partners highlighted that even after its passage, intelligence agencies have systematically surveilled, harassed, and attacked CAJAR members in violation of their rights. 

As we argued, that didn’t happen despite Colombia’s intelligence legal framework, rather it was enabled by its flaws. We emphasized that the Intelligence Law gives authorities wide latitude to surveil human rights defenders, lacking provisions for prior, well-founded, judicial authorization for specific surveillance measures, and robust independent oversight. We also pointed out that Colombian legislation failed to provide the necessary means for defenders to correct and erase their data unlawfully held in intelligence records. 

The court ruled that, as reparation, Colombia must adjust its intelligence legal framework to reflect Inter-American human rights standards. This means that intelligence norms must be changed to clearly establish the legitimate purposes of intelligence actions, the types of individuals and activities subject to intelligence measures, the level of suspicion needed to trigger surveillance by intelligence agencies, and the duration of surveillance measures. 

The reparations also call for Colombia to keep files and records of all steps of intelligence activities, “including the history of access logs to electronic systems, if applicable,” and deliver periodic reports to oversight entities. The legislation must also subject communications surveillance measures to prior judicial authorization, except in emergency situations. Moreover, Colombia needs to pass regulations for mechanisms ensuring the right to informational self-determination in relation to intelligence files. 

These are just some of the fixes the ruling calls for, and they represent a major win. Still, the court missed the opportunity to vehemently condemn state mass surveillance (which can occur under an ill-defined measure in Colombia’s Intelligence Law enabling spectrum monitoring), although Colombian courts will now have the chance to rule it out.

In all, the court ordered the state to take 16 reparation measures, including implementing a system for collecting data on violence against human rights defenders and investigating acts of violence against victims. The government must also publicly acknowledge responsibility for the violations. 

The Inter-American Court's ruling in the CAJAR case sends an important message to Colombia, and the region, that intelligence powers are only lawful and legitimate when there are solid and effective controls and safeguards in place. Intelligence authorities cannot act as if international human rights law doesn't apply to their practices.  

When they do, violations must be fiercely investigated and punished. The ruling elaborates on crucial standards that States must fulfill to make this happen. Only time will tell how closely Colombia and other States will apply the court's findings to their intelligence activities. What’s certain is the dire need to fix a system that helped Colombia become the deadliest country in the Americas for human rights defenders last year, with 70 murders, more than half of all such murders in Latin America. 

EFF and More Than 100+ NGOS Set Non-Negotiable Redlines Ahead of UN Cybercrime Treaty Negotiations

Par : George Wong
23 janvier 2024 à 09:44

EFF has joined forces with 110 NGOs today in a joint statement delivered to the United Nations Ad Hoc Committee, clearly outlining civil society non-negotiable redlines for the proposed UN Cybercrime Treaty, and asserting that states should reject the proposed treaty if these essential changes are not implemented. 

The last draft published on November 6, 2023 does not adequately ensure adherence to human rights law and standards. Initially focused on cybercrime, the proposed Treaty has alarmingly evolved into an expansive surveillance tool.

Katitza Rodriguez, EFF Policy Director for Global Privacy, asserts ahead of the upcoming concluding negotiations:

The proposed treaty needs more than just minor adjustments; it requires a more focused, narrowly defined approach to tackle cybercrime. This change is essential to prevent the treaty from becoming a global surveillance pact rather than a tool for effectively combating core cybercrimes. With its wide-reaching scope and invasive surveillance powers, the current version raises serious concerns about cross-border repression and potential police overreach. Above all, human rights must be the treaty's cornerstone, not an afterthought. If states can't unite on these key points, they must outright reject the treaty.

Historically, cybercrime legislation has been exploited to target journalists and security researchers, suppress dissent and whistleblowers, endanger human rights defenders, limit free expression, and justify unnecessary and disproportionate state surveillance measures. We are concerned that the proposed Treaty, as it stands now, will exacerbate these problems. The proposed treaty concluding session will be held at the UN Headquarters in New York from January 29 to February 10th. EFF will be attending in person.

The joint statement specifically calls States to narrow the scope of criminalization provisions to well defined cyber dependent crimes; shield security researchers, whistleblowers, activists, and journalists from being prosecuted for their legitimate activities; explicitly include language on international human rights law, data protection, and gender mainstreaming; limit the scope of the domestic criminal procedural measures and international cooperation to core cybercrimes established in the criminalization chapter; and address concerns that the current draft could weaken cybersecurity and encryption. Additionally, it requires the necessity to establish specific safeguards, such as the principles of prior judicial authorization, necessity, legitimate aim, and proportionality.

The Growing Threat of Cybercrime Law Abuse: LGBTQ+ Rights in MENA and the UN Cybercrime Draft Convention

This is Part II  of a series examining the proposed UN Cybercrime Treaty in the context of LGBTQ+ communities. Part I looks at the draft Convention’s potential implications for LGBTQ+ rights. Part II provides a closer look at how cybercrime laws might specifically impact the LGBTQ+ community and activists in the Middle East and North Africa (MENA) region.

In the digital age, the rights of the LGBTQ+ community in the Middle East and North Africa (MENA) are gravely threatened by expansive cybercrime and surveillance legislation. This reality leads to systemic suppression of LGBTQ+ identities, compelling individuals to censor themselves for fear of severe reprisal. This looming threat becomes even more pronounced in countries like Iran, where same-sex conduct is punishable by death, and Egypt, where merely raising a rainbow flag can lead to being arrested and tortured.

Enter the proposed UN Cybercrime Convention. If ratified in its present state, the convention might not only bolster certain countries' domestic surveillance powers to probe actions that some nations mislabel as crimes, but it could also strengthen and validate international collaboration grounded in these powers. Such a UN endorsement could establish a perilous precedent, authorizing surveillance measures for acts that are in stark contradiction with international human rights law. Even more concerning, it might tempt certain countries to formulate or increase their restrictive criminal laws, eager to tap into the broader pool of cross-border surveillance cooperation that the proposed convention offers. 

The draft convention, in Article 35, permits each country to define its own crimes under domestic laws when requesting assistance from other nations in cross-border policing and evidence collection. In certain countries, many of these criminal laws might be based on subjective moral judgments that suppress what is considered free expression in other nations, rather than adhering to universally accepted standards.

Indeed, international cooperation is permissible for crimes that carry a penalty of four years of imprisonment or more; there's a concerning move afoot to suggest reducing this threshold to merely three years. This is applicable whether the alleged offense is cyber or not. Such provisions could result in heightened cross-border monitoring and potential repercussions for individuals, leading to torture or even the death penalty in some jurisdictions. 

While some countries may believe they can sidestep these pitfalls by not collaborating with countries that have controversial laws, this confidence may be misplaced. The draft treaty allows countries to refuse a request if the activity in question is not a crime in its domestic regime (the principle of "dual criminality"). However, given the current strain on the MLAT system, there's an increasing likelihood that requests, even from countries with contentious laws, could slip through the checks. This opens the door for nations to inadvertently assist in operations that might contradict global human rights norms. And where countries do share the same subjective values and problematically criminalize the same conduct, this draft treaty seemingly provides a justification for their cooperation.

One of the more recently introduced pieces of legislation that exemplifies these issues is the Cybercrime Law of 2023 in Jordan. Introduced as part of King Abdullah II’s modernization reforms to increase political participation across Jordan, this law was issued hastily and without sufficient examination of its legal aspects, social implications, and impact on human rights. In addition to this new law, the pre-existing cybercrime law in Jordan has already been used against LGBTQ+ people, and this new law expands its capacity to do so. This law, with its overly broad and vaguely defined terms, will severely restrict individual human rights across that country and will become a tool for prosecuting innocent individuals for their online speech. 

Article 13 of the Jordan law expansively criminalizes a wide set of actions tied to online content branded as “pornographic,” from its creation to distribution. The ambiguity in defining what is pornographic could inadvertently suppress content that merely expresses various sexualities, mistakenly deeming them as inappropriate. This goes beyond regulating explicit material; it can suppress genuine expressions of identity. The penalty for such actions entails a period of no less than six months of imprisonment. 

Meanwhile, the nebulous wording in Article 14 of Jordan's laws—terms like “expose public morals,” “debauchery,” and “seduction”—is equally concerning. Such vague language is ripe for misuse, potentially curbing LGBTQ+ content by erroneously associating diverse sexual orientation with immorality. Both articles, in their current form, cast shadows on free expression and are stark reminders that such provisions can lead to over-policing online content that is not harmful at all. During debates on the bill in the Jordanian Parliament, some MPs claimed that the new cybercrime law could be used to criminalize LGBTQ+ individuals and content online. Deputy Leader of the Opposition, Saleh al Armouti, went further and claimed that “Jordan will become a big jail.” 

Additionally, the law imposes restrictions on encryption and anonymity in digital communications, preventing individuals from safeguarding their rights to freedom of expression and privacy. Article 12 of the Cybercrime Law prohibits the use of Virtual Private Networks (VPNs) and other proxies, with at least six months imprisonment or a fine for violations. 

This will force people in Jordan to choose between engaging in free online expression or keeping their personal identity private. More specifically, this will negatively impact LGBTQ+ people and human rights defenders in Jordan who particularly rely on VPNs and anonymity to protect themselves online. The impact of Article 12 is exacerbated by the fact that there is no comprehensive data privacy legislation in Jordan to protect people’s rights during cyber attacks and data breaches.  

This is not the first time Jordan has limited access to information and content online. In December 2022, Jordanian authorities blocked TikTok to prevent the dissemination of live updates and information during the workers’ protests in the country's south, and authorities there previously had blocked Clubhouse as well

This crackdown on free speech has particularly impacted journalists, such as the recent arrest of Jordanian journalist Heba Abu Taha for criticizing Jordan’s King over his connections with Israel. Given that online platforms like TikTok and Twitter are essential for activists, organizers, journalists, and everyday people around the world to speak truth to power and fight for social justice, the restrictions placed on free speech by Jordan’s new Cybercrime Law will have a detrimental impact on political activism and community building across Jordan.

People across Jordan have protested the law and the European Union has  expressed concern about how the law could limit freedom of expression online and offline. In August, EFF and 18 other civil society organizations wrote to the King of Jordan, calling for the rejection of the country’s draft cybercrime legislation. With the law now in effect, we urge Jordan to repeal the Cybercrime Law 2023.

Jordan’s Cybercrime Law has been said to be a “true copy” of the United Arab Emirates (UAE) Federal Decree Law No. 34 of 2021 on Combatting Rumors and Cybercrimes. This law replaced its predecessor, which had been used to stifle expression critical of the government or its policies—and was used to sentence human rights defender Ahmed Mansoor to 10 years in prison. 

The UAE’s new cybercrime law further restricts the already heavily-monitored online space and makes it harder for ordinary citizens, as well as journalists and activists, to share information online. More specifically, Article 22 mandates prison sentences of between three and 15 years for those who use the internet to share “information not authorized for publishing or circulating liable to harm state interests or damage its reputation, stature, or status.” 

In September 2022, Tunisia passed its new cybercrime law in Decree-Law No. 54 on “combating offenses relating to information and communication systems.” The wide-ranging decree has been used to stifle opposition free speech, and mandates a five-year prison sentence and a fine for the dissemination of “false news” or information that harms “public security.” In the year since Decree-Law 54 was enacted, authorities in Tunisia have prosecuted media outlets and individuals for their opposition to government policies or officials. 

The first criminal investigation under Decree-Law 54 saw the arrest of student Ahmed Hamada in October 2022 for operating a Facebook page that reported on clashes between law enforcement and residents of a neighborhood in Tunisia. 

Similar tactics are being used in Egypt, where the 2018 cybercrime law, Law No. 175/2018, contains broad and vague provisions to silence dissent, restrict privacy rights, and target LGBTQ+ individuals. More specifically, Articles 25 and 26 have been used by the authorities to crackdown on content that allegedly violates “family values.” 

Since its enactment, these provisions have also been used to target LGBTQ+ individuals across Egypt, particularly regarding the publication or sending of pornography under Article 8, as well as illegal access to an information network under Article 3. For example, in March 2022 a court in Egypt charged singers Omar Kamal and Hamo Beeka with “violating family values” for dancing and singing in a video uploaded to YouTube. In another example, police have used cybercrime laws to prosecute LGBTQ+ individuals for using dating apps such as Grindr.

And in Saudi Arabia, national authorities have used cybercrime regulations and counterterrorism legislation to prosecute online activism and stifle dissenting opinions. Between 2011 and 2015, at least 39 individuals were jailed under the pretense of counterterrorism for expressing themselves online—for composing a tweet, liking a Facebook post, or writing a blog post. And while Saudi Arabia has no specific law concerning gender identity and sexual orientation, authorities have used the 2007 Anti-Cyber Crime Law to criminalize online content and activity that is considered to impinge on “public order, religious values, public morals, and privacy.” 

These provisions have been used to prosecute individuals for peaceful actions, particularly since the Arab Spring in 2011. More recently, in August 2022, Salma al-Shehab was sentenced to 34 years in prison with a subsequent 34-year travel ban for her alleged “crime” of sharing content in support of prisoners of conscience and women human rights defenders.

These cybercrime laws demonstrate that if the proposed UN Cybercrime Convention is ratified in its current form with its broad scope, it would authorize domestic surveillance for the investigation of any offenses, as those in Articles 12, 13, and 14 of Jordan's law. Additionally, the convention could authorize international cooperation for investigation of crimes penalized with three or four years of imprisonment, as seen in countries such as the UAE, Tunisia, Egypt, and Saudi Arabia.

As Canada warned (at minute 01:56 ) at the recent negotiation session, these expansive provisions in the Convention permit states to unilaterally define and broaden the scope of criminal conduct, potentially paving the way for abuse and transnational repression. While the Convention may incorporate some procedural safeguards, its far-reaching scope raises profound questions about its compatibility with the key tenets of human rights law and the principles enshrined in the UN Charter. 

The root problem lies not in the severity of penalties, but in the fact that some countries criminalize behaviors and expression that are protected under international human rights law and the UN Charter. This is alarming, given that numerous laws affecting the LGBTQ+ community carry penalties within these ranges, making the potential for misuse of such cooperation profound.

In a nutshell, the proposed UN treaty amplifies the existing threats to the LGBTQ+ community. It endorses a framework where nations can surveil benign activities such as sharing LGBTQ+ content, potentially intensifying the already-precarious situation for this community in many regions.

Online, the lack of legal protection of subscriber data threatens the anonymity of the community, making them vulnerable to identification and subsequent persecution. The mere act of engaging in virtual communities, sharing personal anecdotes, or openly expressing relationships could lead to their identities being disclosed, putting them at significant risk.

Offline, the implications intensify with amplified hesitancy to participate in public events, showcase LGBTQ+ symbols, or even undertake daily routines that risk revealing their identity. The draft convention's potential to bolster digital surveillance capabilities means that even private communications, like discussions about same-sex relationships or plans for LGBTQ+ gatherings, could be intercepted and turned against them. 

To all member states: This is a pivotal moment. This is our opportunity to ensure the digital future is one where rights are championed, not compromised. Pledge to protect the rights of all, especially those communities like the LGBTQ+ that are most vulnerable. The international community must unite in its commitment to ensure that the proposed convention serves as an instrument of protection, not persecution.



❌
❌