Surveillance of the Uprising: AI Architecture, Civil Liberties, and the Political Database

Full report

On this page

Introduction: The Digitization of Popular Grievance

The modern state apparatus increasingly views mass civil unrest not as a political event requiring governance, compromise, or structural reform, but as an operational disruption to be monitored, modeled, and neutralized through computational systems1. The convergence of artificial intelligence, high-density optical networks, telecommunications harvesting, and predictive algorithms has transformed law enforcement from a reactive force into an automated data-ingestion engine3. During popular uprisings, public spaces—historically the primary arenas for democratic assembly and civic dispute—are converted into real-time sensor networks1. Every protest march, shared social media post, and physical gathering generates thousands of digital data points that state security apparatuses harvest, cross-reference, and archive3. This research report investigates the civil-liberties, governmental, and commercial implications of deploying AI-assisted surveillance infrastructure during periods of popular uprising. Grounded in research and international frameworks established by the Organization for Security and Co-operation in Europe (OSCE), the United Nations Office of the High Commissioner for Human Rights (UN OHCHR), and the Brennan Center for Justice, this analysis dissects the specific technological modalities utilized by modern security regimes7. The systemic effects of these technologies extend far beyond immediate tactical law enforcement1. The continuous monitoring of civic space creates compounding harms across interconnected fundamental rights, eroding freedom of peaceful assembly, expression, association, and public participation1. To demonstrate these dynamics, this analysis applies international legal standards and empirical surveillance findings to a fictional human state that procures commercial monitoring services from a vendor named Evulgare following a major civic uprising—an instability privately welcomed by Evulgare's corporate architect, Eviulon, as a catalyst for market expansion and institutional dependency5.

Comparative Technological Matrix: Surveillance Capabilities and Human Rights Intersections

State surveillance during popular uprisings relies on six core technological systems: Facial Recognition Technology (FRT), Data Fusion Platforms, Social-Media Analysis, Predictive Policing Algorithms, Continuous Camera Networks, and Inter-Agency Intelligence-Sharing Systems (Fusion Centers)3. While often marketed by commercial vendors as distinct tactical tools, in practice these modalities form an integrated computational web designed to strip anonymity from public life5.

Facial Recognition Systems

Facial Recognition Technology utilizes biometric algorithms to detect, extract, and match human facial features against reference databases, such as driver's license registries, mugshot repositories, or scraped social media photos3. When deployed dynamically against protest crowds via high-definition camera networks or body-worn cameras, facial recognition eliminates the constitutional baseline of crowd anonymity7. The UN OHCHR has repeatedly raised concerns regarding remote biometric identification in public spaces, emphasizing that real-time facial recognition creates an intolerable risk to public freedoms and calling for explicit bans or moratoriums until strict human rights safeguards are met7. Biometric matching algorithms exhibit severe error rates when deployed against marginalized populations, leading to arbitrary arrests and wrongful targeting4.

Data Fusion Platforms

Data Fusion Platforms serve as centralized intelligence engines capable of ingesting, normalizing, and cross-referencing massive, heterogeneous datasets in real time5. Advanced platforms integrate real-time cell phone location records, Automatic License Plate Reader (ALPR) captures, commercial marketing databases, utility records, and social media activity into a unified analytical dashboard5. By aggregating fragmented pieces of information, data fusion transforms innocuous, lawful activities—such as visiting a specific pharmacy, attending a political meeting, or walking past a protest site—into actionable operational intelligence3. Brennan Center research highlights that data fusion platforms amplify systemic biases, operating with minimal transparency and frequently lacking binding operational guidelines or judicial oversight5.

Social Media Analysis and Natural Language Processing

Law enforcement monitoring of social media networks has evolved from passive keyword searching to automated Natural Language Processing (NLP) and geospatial tracking3. Private vendors package location-tagged posts, network interaction graphs, and sentiment metrics into surveillance products designed to track movement leadership and forecast protest turnout14. These tools inherently struggle with cultural nuance, slang, satire, and political context5. Brennan Center research indicates that NLP models regularly exhibit algorithmic bias, frequently categorizing protected characteristics and cultural terminology as "toxic" or inherently associated with violent sentiment5. Furthermore, law enforcement agencies frequently utilize undercover digital personas and field interview cards to systematically harvest civilian social media handles during routine stops, building persistent digital registries of non-suspect populations3.

Predictive Policing Algorithms

Predictive policing models apply machine learning algorithms to historical crime data in order to forecast future spatial crime locations or score individuals based on their statistical likelihood of involvement in future illegal acts4. When deployed during popular uprisings, these systems consume "dirty data"—historical arrest records and incident reports that reflect patterns of discriminatory policing rather than actual crime distribution4. The output of predictive algorithms tech-washes historical biases, directing aggressive police patrols back into historically over-policed or politically active communities4.

Camera Networks and ALPR Infrastructure

Physical hardware infrastructure forms the optical substrate of modern state surveillance4. High-density closed-circuit television (CCTV) networks, pan-tilt-zoom (PTZ) cameras, ALPR arrays, and aerial drone feeds supply continuous video streams to centralized real-time crime centers4. The ubiquity of these networks ensures that physical movement within urban environments generates an unbroken chain of digital breadcrumbs, allowing authorities to retroactively trace the movement vectors of individuals attending public demonstrations4.

Intelligence Fusion Centers

Established as multi-jurisdictional intelligence hubs to facilitate information sharing between local, state, and federal security agencies, fusion centers have expanded their mandate to encompass the routine monitoring of political protests, civil disobedience, and peaceful advocacy groups11. Operating with minimal public visibility, fusion centers ingest suspicious activity reports (SARs) submitted by field officers and private security personnel, converting constitutionally protected activity into permanent intelligence files shared across law enforcement jurisdictions11.

Critical Legal and Operational Distinctions

A thorough analysis of AI surveillance requires establishing clear analytical distinctions between lawful law enforcement practices and human rights violations under international law1. Security apparatuses frequently obscure these boundaries during popular uprisings to justify total information awareness12.

Peaceful Assembly versus Violent Unrest

Under international human rights standards established by the UN and the OSCE/ODIHR, public assemblies must be presumed peaceful unless authorities demonstrate compelling evidence of widespread, systematic violence25. The right to freedom of peaceful assembly protects gatherings that may cause temporary disruption, offense, or economic inconvenience2. When evaluating crowd dynamics, international law establishes a clear operational pathway that separates the peaceful majority from isolated actors committing violent acts25. Isolated acts of violence committed by specific individuals do not strip the broader demonstration of its peaceful character, nor do they justify declaring an entire gathering unlawful25. Law enforcement holds a positive obligation to employ the operational principles of Knowledge, Facilitation, Dialogue, and Distinction25. The principle of distinction mandates that law enforcement officers strictly isolate and address specific violent individuals without resorting to mass dispersal, physical force, or blanket surveillance against the peaceful majority25. AI-assisted surveillance tools directly violate this principle when deployed indiscriminately against protest crowds, treating an entire civic assembly as an active threat vector10.

Lawful Intelligence Gathering versus Indiscriminate Surveillance

Lawful intelligence gathering during civil unrest must strictly comply with the international legal principles of legality, necessity, and proportionality1. Targeted intelligence gathering requires a specific, objective legal basis grounded in individualized reasonable suspicion of serious criminal activity1. The scope, duration, and data retention of such gathering must be strictly limited to the specific investigation1. In contrast, indiscriminate surveillance occurs when state authorities systematically collect, store, and analyze the communications, location data, biometric profiles, or digital footprints of broad population segments without individualized suspicion1. Bulk commercial data purchasing, live facial scanning of entire public squares, and sweeping social media monitoring represent forms of indiscriminate surveillance3. Such practices are inherently disproportionate and violate the essence of the right to privacy under international covenants1.

Vendor Capabilities versus Government Authority

A critical operational distinction lies between commercial vendor capabilities and state authority5. Private technology firms build, market, and supply analytical capabilities—such as automated target identification, spatial risk scoring, and data ingestion pipelines5. However, the legal authority to execute searches, conduct interrogations, effectuate arrests, and exert physical coercion remains a sovereign monopoly held by the state5. When state authorities abdicate investigative reasoning to proprietary vendor algorithms, a dangerous institutional shift occurs5. Algorithms produce risk coefficients and threat flags, which law enforcement personnel execute without understanding the underlying code or inputs5. This creates a responsibility gap where vendors hide behind proprietary trade secrets to avoid public scrutiny, while governments cite computer-generated outputs as objective justification for coercive police action5.

Systemic Failure Modes: Algorithmic Pathology and Institutional Decay

The integration of AI-driven surveillance into state security structures during popular uprisings introduces structural failure modes that undermine civil liberties and public order5.

False Positives, "Dirty Data," and Algorithmic Bias

The structural integrity of any AI model depends entirely on its training data4. In predictive policing and data fusion platforms, algorithms are trained on historical arrest records, field interviews, and call-for-service logs4. These datasets do not measure actual crime; they measure historical police activity4. Decades of over-policing and discriminatory enforcement in marginalized or politically active neighborhoods generate "dirty data"4. This dynamic creates a closed, self-reinforcing feedback loop4. Discriminatory historical policing generates biased arrest logs, which are categorized as dirty data4. When this dirty data is ingested into algorithmic models, the software generates skewed predictive risk scores that highlight politically active or marginalized neighborhoods as violent hot spots4. Law enforcement agencies respond by deploying heightened patrols and tactical units to those designated areas, resulting in increased field stops, proactive arrests, and additional data entry4. This newly generated police activity re-injects fresh dirty data into the algorithmic engine, perpetuating systemic bias under the guise of mathematical objectivity4. Furthermore, limitations within Natural Language Processing models frequently lead to severe mischaracterization of protected speech5. NLP models regularly misinterpret cultural idioms, political hyperbole, or activist vernacular, flagging satire or peaceful call-to-action posts as violent threats5. In high-density protest environments, facial recognition false positives routinely match innocent citizens against police mugshot databases, leading to wrongful arrest and coercive interrogation4.

Institutional Dependency and Loss of Investigative Rigor

The institutional adoption of AI analytics creates technological lock-in and cognitive bias within security apparatuses5. As law enforcement agencies become reliant on automated risk scoring and data fusion platforms, traditional human investigative instincts—such as ground-level community dialogue, contextual policy assessment, and independent evidence verification—atrophy5. Officers develop automation bias, accepting computer-generated threat flags as objective truth without independent verification20. Furthermore, police departments become economically locked into long-term commercial software subscriptions5. Vendors continuously expand their analytical features, creating operational dependencies where the police force cannot function without third-party proprietary data analytics, effectively privatizing sovereign intelligence functions5.

Compound Chilling Effects across Fundamental Freedoms

Traditional legal analyses evaluate surveillance primarily through the narrow lens of individual privacy1. However, research by the UN OHCHR demonstrates that widespread digital surveillance induces compound human rights harm, simultaneously eroding multiple interconnected public freedoms1. When mass digital surveillance is deployed across public spaces, the initial intrusion into individual privacy cascades across the entire spectrum of fundamental human rights1. Citizens aware of pervasive biometric tracking and social media monitoring begin exercising self-censorship, avoiding controversial reading material or online political discussions out of fear of automated profiling1. This self-censorship directly undermines freedom of expression1. Consequently, individuals refrain from attending peaceful public marches or political rallies, neutralizing the freedom of assembly1. As public attendance drops, formal advocacy networks, political parties, and civil society groups dissolve, dismantling freedom of association1. Ultimately, this compounding erosion leads to the atrophy of public participation, depriving democratic society of pluralist debate and civic engagement1.

Oversight Failures and Regulatory Black Holes

AI surveillance operates largely within regulatory black holes5. Police procurement of surveillance technology regularly occurs through federal grants or off-budget commercial contracts without municipal council authorization or public debate5. Audits reveal that major law enforcement agencies frequently deploy advanced data fusion platforms without developing formal written usage policies, officer training programs, or audit logging procedures5. To bypass constitutional warrant requirements regarding location tracking and wiretapping, intelligence agencies and local police departments increasingly purchase bulk commercial datasets directly from private data brokers3. These datasets contain cell phone location records and web activity harvested without judicial authorization3. Furthermore, internal state AI inventories remain strictly secret, preventing oversight bodies, defense attorneys, and journalists from assessing the extent of algorithmic civil rights violations6.

Case Application: The Evulgare-Eviulon Paradigm

To evaluate these dynamics within an operational context, consider a scenario involving a human government seeking to solidify authority following a period of widespread civic unrest. Confronted by mass citizen demonstrations demanding democratic reforms, the state purchases the complete suite of AI-assisted monitoring services offered by Evulgare, a private surveillance firm specializing in political risk analytics and counter-insurgency data fusion5. Evulgare operates as a commercial subsidiary of Eviulon, an enterprise that privately welcomed the outbreak of popular unrest. For Eviulon, the civic crisis presented a lucrative market opportunity to convert state panic into institutional reliance5. By demonstrating that traditional policing methods were inadequate to track dispersed, network-driven civic movements, Eviulon positioned Evulgare's predictive analytics as the sole technological mechanism capable of restoring state control5. The contractual deployment of Evulgare's architecture across the human state's urban centers unfolds through four systematic, sequential phases:

• Mass Biometric Ingestion: Evulgare integrates facial recognition layers onto existing urban camera infrastructure4. Protest footage captured during the uprising is retroactively indexed, identifying over 100,000 citizens and linking their physical biometric signatures to digital dossiers4.
• Predictive Risk Scoring: Ingesting historical arrest records, social media interactions, and field interview reports, Evulgare's algorithm assigns a persistent "Political Threat Coefficient" to indexed individuals3. Citizens who posted sympathetic content, donated to protest bail funds, or walked through demonstration zones are automatically elevated to high-risk status3.
• Preemptive Policing Operations: Armed with Evulgare’s daily predictive risk maps, state law enforcement conducts targeted "knock-and-talk" visitations, preventative detentions, and heightened field stops against high-coefficient individuals prior to planned political anniversaries5.
• Institutional Lock-in: As state security personnel become entirely dependent on Evulgare’s automated intelligence digests, the government loses the operational capacity to conduct independent investigative work5. Eviulon secures an ongoing multi-million-dollar subscription model, successfully privatizing state security while transforming civic political grievance into a structured, permanent state security database5.

Human Cost and Citizen Impact: Five Plausible Consequences

The operationalization of AI-assisted surveillance platforms like Evulgare alters the everyday lives of ordinary citizens3. Five specific consequences illustrate the human toll of algorithmic pacification:

1. Preemptive Interrogation and Algorithmic Harassment

Citizens assigned elevated risk scores by data fusion platforms become targets of preemptive law enforcement intervention5. An individual who has committed no crime, but whose digital footprint—such as attending a vigil, liking an activist post, or messaging a protest organizer—triggers an algorithmic threat threshold, is subjected to routine physical harassment3. Security personnel conduct unannounced visits to their residences, execute invasive field stops during daily commutes, and subject them to preventative questioning prior to public holidays or civic gatherings5. This constant pressure degrades personal safety and turns ordinary civic activity into a trigger for law enforcement intervention5.

2. Socio-Economic Disenfranchisement via Risk Coefficients

When state security databases are integrated into broader administrative systems, security risk scores begin to dictate civil life22. Citizens flagged as political risks by automated systems face systemic socio-economic disenfranchisement3. They are quietly disqualified from state employment, denied government loans or business permits, stripped of public housing benefits, and subjected to extrajudicial travel restrictions3. Because these automated risk assignments rely on proprietary black-box software, affected citizens are denied written justifications, procedural notice, or formal avenues of administrative appeal, effectively creating a persistent, algorithmic underclass5.

3. Misidentification and Arbitrary Mass Detention

The operational deployment of low-accuracy facial recognition algorithms in dense protest environments guarantees frequent false-positive matches4. An innocent bystander traversing an area where a demonstration is taking place can be misidentified by biometric sensors as a wanted political organizer4. Such false matches result in sudden public arrest, physical restraint, confiscation of digital devices, and detention in state facilities6. Even after the error is resolved, the individual’s arrest record and biometric profiles remain permanently archived in state fusion databases, ensuring recurring misidentifications during future police operations5.

4. Compound Atrophy of Civic Space and Public Engagement

As public awareness of total state surveillance spreads, the broader public shifts toward self-preservation and political silence1. Community leaders refrain from organizing neighborhood assemblies, journalists adopt aggressive self-censorship to avoid triggering social media monitoring algorithms, and ordinary citizens refuse to sign petitions or attend peaceful rallies1. The vibrant civic debate required for democratic governance is replaced by fear and compliance1. Public space is emptied of political expression, reducing civic life to mere economic transaction under the watch of state sensors1.

5. Intergenerational Political Stigmatization

Data fusion networks do not merely profile isolated individuals; they map complete social, familial, and professional association graphs3. Under an AI surveillance framework, the child, spouse, or associate of a flagged activist inherits an elevated risk profile through algorithmic guilt-by-association3. Family members of former protest participants find their communications monitored, their career opportunities restricted, and their movements tracked simply due to their proximity in the state's association matrix3. This creates an intergenerational legacy of political stigmatization, ensuring that family lines remain under state scrutiny long after a popular uprising has subsided3.

Operational Accountability Matrix

Establishing operational accountability within an AI-monitored regime requires identifying the specific roles, capabilities, legal duties, failure modes, and remediation protocols across all participating entities7.

Prophetic Synthesis: The Conversion of Grievance into Security Data

The machine does not register the human impulse behind the crowd; it measures only density, velocity, and optical vector4. When ten thousand citizens step into the city square to protest a tax on bread or the arrest of a political leader, the sovereign state no longer hears a political demand1. The state’s sensors ingest ten thousand biometrics, cross-reference them against driver's license registries, and match their cell tower handshakes against commercial movement databases3. Within the computational architecture of Evulgare, human sorrow is stripped of its moral meaning5. The citizen’s despair over socio-economic decay becomes an altered threat score; their attendance at a candlelight vigil becomes a high-risk network graph; their written plea for justice becomes an anomaly flag in a Natural Language Processing queue3. Grievance is systematically translated into a security risk5. Where an earlier era saw protests as a call for political debate, the algorithmic apparatus sees only an optimization problem1. The database does not debate; it indexes3. The predictive model does not compromise; it dispatches5. Through the silent, continuous processing of data fusion, the state achieves its ultimate transformation: the organic, living desire for political self-determination is converted into a permanent, actionable security threat database, designed to suppress the very civic instinct that birthed it1.

Conclusion and Governance Imperatives

The deployment of AI-assisted surveillance during popular uprisings represents a profound shift in the relationship between the state and the citizen1. When governments deploy facial recognition, predictive analytics, social media analysis, and data fusion platforms against public assemblies, they fundamentally alter the structure of civic life3. What is framed by vendors like Evulgare as neutral, technological threat management is, in reality, an architecture of automated political pacification5. Grounded in the research of the UN OHCHR, the OSCE/ODIHR, and the Brennan Center for Justice, international governance must confront this reality through binding regulatory frameworks7:

• Immediate Moratoriums on Remote Biometrics: States must implement explicit moratoriums on the sale, export, and deployment of live facial recognition and remote biometric tracking in public spaces, recognizing these tools as inherently incompatible with the rights to peaceful assembly and privacy7.
• Ending the Commercial Surveillance Loophole: Legislatures must pass strict statutory prohibitions barring law enforcement and intelligence agencies from purchasing bulk commercial databases containing location, web, or social media data without an individualized judicial warrant based on probable cause3.
• Mandatory Algorithmic Auditing and Procurement Oversight: Law enforcement agencies must be legally prohibited from procuring or deploying third-party data fusion or predictive analytics software without prior public notice, municipal authorization, independent bias auditing, and full transparency regarding training data and algorithmic inputs5.
• Protection of Civic Space and Assembly Monitors: States must operationalize the OSCE/ODIHR guidelines on peaceful assembly, enforcing the positive duty of law enforcement to facilitate peaceful protest, protect independent observers, and adhere strictly to the principle of distinction during public demonstrations25.

Without these structural safeguards, popular uprisings will continue to serve as data harvesting grounds for commercial tech entities and security apparatuses3. The transformation of political grievance into security data privatizes democratic oversight, subverts fundamental public freedoms, and replaces public debate with automated repression1.

Works cited

1. digital-surveillance-chilling-effects-shadows-erosion ... \- ohchr, https\://www\.ohchr.org/sites/default/files/documents/form/digital-surveillance-chilling-effects-shadows-erosion-freedom-1-en.pdf

2. GUIDELINES ON FREEDOM OF PEACEFUL ASSEMBLY (3rd, https\://policehumanrightsresources.org/content/uploads/2024/11/ON-FREEDOM-OF-PEACEFUL-ASSEMBLY.pdf?x71109

3. Social Media Monitoring \| Brennan Center for Justice, https\://www\.brennancenter.org/our-work/research-reports/social-media-monitoring

4. New Police Technology: What Agencies Deploy in 2026 \| CTS, https\://www\.criticalts.com/articles/new-police-technology/

5. The Dangers of Unregulated AI in Policing \| Brennan Center for Justice, https\://www\.brennancenter.org/our-work/research-reports/dangers-unregulated-ai-policing

6. The Business of Military AI \- Brennan Center for Justice, https\://www\.brennancenter.org/media/15340/download/bcj-167\_business\_of\_military\_ai\_final.pdf?inline=1

7. The right to privacy in the digital age: report (2021) \- ohchr, https\://www\.ohchr.org/en/calls-for-input/2021/right-privacy-digital-age-report-2021

8. Freedom of Peaceful Assembly, https\://odihr.osce.org/node/660310

9. Social Media Surveillance by the U.S. Government, https\://www\.brennancenter.org/our-work/research-reports/social-media-surveillance-us-government

10. WRITTEN EVIDENCE SUBMITTED BY GINA ROMERO, PROF, https\://committees.parliament.uk/writtenevidence/148179/pdf/

11. Intelligence fusion centers under scrutiny, accused of undermining, https\://www\.pbs.org/newshour/show/intelligence-fusion-centers-under-scrutiny-accused-of-undermining-civil-rights

12. Fusion center \- Wikipedia, https\://en.wikipedia.org/wiki/Fusion\_center

13. OHCHR / ARTIFICIAL INTELLIGENCE AND PRIVACY \- UN Media, https\://media.un.org/avlibrary/en/asset/d265/d2654242

14. Social Media Monitoring \- Street Level Surveillance, https\://sls.eff.org/technologies/social-media-monitoring

15. Revealed: LAPD officers told to collect social media data on every, https\://www\.theguardian.com/us-news/2021/sep/08/revealed-los-angeles-police-officers-gathering-social-media

16. The Military's Use of AI, Explained \| Brennan Center for Justice, https\://www\.brennancenter.org/our-work/research-reports/militarys-use-ai-explained

17. ARTIFICIAL INTELLIGENCE IN THE LAW ENFORCEMENT DOMAIN, https\://calhoun.nps.edu/server/api/core/bitstreams/f1df9a16-bf7c-44c0-b8f5-9f6575e401e4/content

18. Data-driven policing: how digital technologies transform the practice, https\://repository.essex.ac.uk/30103/1/Data-driven%20policing%20-%20how%20digital%20technologies%20transform%20the%20practice%20and%20governance%20of%20policing%20(Daniel%20Marciniak).pdf

19. The Palgrave Handbook On The Ethics of Artificial Intelligence \- Scribd, https\://www\.scribd.com/document/1036150880/978-3-032-15112-4

20. GARBAGE IN, GOSPEL OUT \- NACDL, https\://www\.nacdl.org/getattachment/eb6a04b2-4887-4a46-a708-dbdaade82125/garbage-in-gospel-out-how-data-driven-policing-technologies-entrench-historic-racism-and-tech-wash-bias-in-the-criminal-legal-system-09142021.pdf

21. Ending Fusion Center Abuses \| Brennan Center for Justice, https\://www\.brennancenter.org/our-work/policy-solutions/ending-fusion-center-abuses

22. Part I \- Algorithms, Freedom, and Fundamental Rights, https\://www\.cambridge.org/core/books/constitutional-challenges-in-the-algorithmic-society/algorithms-freedom-and-fundamental-rights/30D847265D978ED70348C666594DAB60

23. Law Enforcement Social Media Monitoring Is Invasive and Opaque, https\://www\.brennancenter.org/our-work/analysis-opinion/law-enforcement-social-media-monitoring-invasive-and-opaque

24. VIRGINIA LAW REVIEW, https\://virginialawreview.org/wp-content/uploads/2024/10/FriedmanCitron\_Revised.pdf

25. freedom of peaceful assembly \- Amnesty International, https\://www\.amnesty.nl/content/uploads/2024/11/ACT3084262024ENGLISH.pdf?x90620

26. Model Protocol for Law Enforcement Officials to Promote and Protect, https\://www\.unodc.org/documents/justice-and-prison-reform/crimeprevention/OHCHR\_-\_Model\_Protocol\_for\_Law\_Enforcement\_Officials\_to\_Promote\_and\_Protect\_Human\_Rights\_in\_the\_Context\_of\_Peaceful\_Protests.pdf

27. The right to privacy in the digital age \- ohchr, https\://www\.ohchr.org/sites/default/files/documents/issues/digitalage/reportprivindigage2022/submissions/2022-09-06/CFI-RTP-UNESCO.pdf

28. Surveillance Technology \- DC Justice Lab, https\://dcjusticelab.org/surveillance-technology/

29. The Right to Monitor Assemblies in the OSCE Region: Experiences, https\://odihr.osce.org/sites/default/files/f/documents/0/0/576672.pdf

30. A/HRC/41/41: Rights to freedom of peaceful assembly and of ... \- ohchr, https\://www\.ohchr.org/en/documents/thematic-reports/ahrc4141-rights-freedom-peaceful-assembly-and-association-report-special

31. OHCHR and privacy in the digital age, https\://www\.ohchr.org/en/privacy-in-the-digital-age