\documentclass{article} % For LaTeX2e
\usepackage{iclr2024_conference,times}
\usepackage{xcolor}

\usepackage[utf8]{inputenc} % allow utf-8 input
\usepackage[T1]{fontenc}    % use 8-bit T1 fonts
\usepackage{hyperref}       % hyperlinks
\usepackage{url}            % simple URL typesetting
\usepackage{booktabs}       % professional-quality tables
\usepackage{amsfonts}       % blackboard math symbols
\usepackage{nicefrac}       % compact symbols for 1/2, etc.
\usepackage{microtype}      % microtypography
\usepackage{titletoc}

\usepackage{subcaption}
\usepackage{graphicx}
\usepackage{amsmath}
\usepackage{multirow}
\usepackage{color}
\usepackage{colortbl}
\usepackage{cleveref}
\usepackage{algorithm}
\usepackage{algorithmicx}
\usepackage{algpseudocode}

\DeclareMathOperator*{\argmin}{arg\,min}
\DeclareMathOperator*{\argmax}{arg\,max}

\graphicspath{{../}} % To reference your generated figures, see below.
\begin{filecontents}{references.bib}
@article{Ballis2022,
  author = {Ballis, Philipp and Schwendemann, Anna},
  title = {Witnessing Digital Suffering: Trust, Credibility, and Moral Authority in Networked Testimonies},
  journal = {Digital Journalism},
  year = {2022},
  volume = {10},
  number = {6},
  pages = {733--759},
  doi = {10.1080/21670811.2022.2039084}
}
@misc{TechForPalestine2025,
  author = {{Tech for Palestine Collective}},
  title = {Boycott Israeli Tech Companies Dataset},
  year = {2025},
  howpublished = {\url{https://github.com/TechForPalestine/boycott-israeli-tech-companies-dataset/tree/main}},
  note = {Accessed October 2025}
}
@report{OHCHR2025,
  author = {{Office of the High Commissioner for Human Rights}},
  title = {Database of Business Enterprises Involved in Israeli Settlements (2025 Update)},
  institution = {United Nations Human Rights Council},
  year = {2025},
  note = {A/HRC/58/43, released 26 September 2025}
}
@report{ICJ2024,
  author = {{International Court of Justice}},
  title = {Application of the Convention on the Prevention and Punishment of the Crime of Genocide in the Gaza Strip (South Africa v. Israel): Provisional Measures},
  year = {2024},
  month = {January},
  note = {Order of 26 January 2024, The Hague}
}
@report{UNSR2024,
  author = {{United Nations Special Rapporteur on the situation of human rights in the Palestinian territories occupied since 1967}},
  title = {Anatomy of a Genocide: Report A/HRC/55/73},
  year = {2024},
  institution = {United Nations Human Rights Council}
}
@report{HRW2024,
  author = {{Human Rights Watch}},
  title = {Gaza: Israel’s Digital Targeting Tools Risk Civilian Harm},
  year = {2024},
  month = {September},
  howpublished = {\url{https://www.hrw.org/news/2024/09/10/gaza-israeli-militarys-digital-tools-risk-civilian-harm}}
}
@report{Amnesty2023,
  author = {{Amnesty International}},
  title = {Automated Apartheid: Facial Recognition Entrenching the Israeli Occupation},
  year = {2023},
  institution = {Amnesty International},
  howpublished = {\url{https://www.amnesty.org/en/documents/mde15/6701/2023/en/}}
}
@report{Amnesty2021,
  author = {{Amnesty International}},
  title = {Forensic Methodology Report: How to Catch NSO Group’s Pegasus},
  year = {2021},
  institution = {Amnesty Tech},
  howpublished = {\url{https://www.amnesty.org/en/latest/research/2021/07/forensic-methodology-report-how-to-catch-nso-groups-pegasus/}}
}
@misc{WhoProfits2024,
  author = {{Who Profits Research Center}},
  title = {Company Profiles: Israeli and International Corporations Involved in the Occupation},
  year = {2024},
  howpublished = {\url{https://www.whoprofits.org/}}
}
@article{Zelizer2021,
  author = {Zelizer, Barbie},
  title = {Why Journalism Is about More than News: Refiguring Practice, Place, and Politics},
  journal = {Communication Theory},
  year = {2021},
  volume = {31},
  number = {4},
  pages = {571--590}
}
@book{Pantti2022,
  author = {Pantti, Mervi},
  title = {Digital Witnessing: A New Era of Testimony},
  publisher = {Oxford University Press},
  year = {2022}
}
@book{Zuboff2019,
  author = {Zuboff, Shoshana},
  title = {The Age of Surveillance Capitalism},
  publisher = {PublicAffairs},
  address = {New York},
  year = {2019}
}
@book{Deibert2021,
  author = {Deibert, Ronald},
  title = {Reset: Reclaiming the Internet for Civil Society},
  publisher = {House of Anansi},
  year = {2021}
}
@book{Fricker2007,
  author = {Fricker, Miranda},
  title = {Epistemic Injustice: Power and the Ethics of Knowing},
  publisher = {Oxford University Press},
  year = {2007}
}
@book{Margalit2002,
  author = {Margalit, Avishai},
  title = {The Ethics of Memory},
  publisher = {Harvard University Press},
  year = {2002}
}
@book{Ruggie2011,
  author = {Ruggie, John Gerard},
  title = {Just Business: Multinational Corporations and Human Rights},
  publisher = {W. W. Norton},
  year = {2011}
}
@book{Ratner2019,
  author = {Ratner, Steven},
  title = {The Thin Justice of International Law},
  publisher = {Oxford University Press},
  year = {2019}
}
@book{Creswell2018,
  author = {Creswell, John W. and Plano Clark, Vicki L.},
  title = {Designing and Conducting Mixed Methods Research},
  publisher = {Sage},
  edition = {3},
  year = {2018}
}
@book{Flick2014,
  author = {Flick, Uwe},
  title = {An Introduction to Qualitative Research},
  publisher = {Sage},
  edition = {5},
  year = {2014}
}
@book{Allan2017,
  author = {Allan, Stuart},
  title = {Citizen Witnessing: Revisioning Journalism in Times of Crisis},
  publisher = {Polity},
  year = {2017}
}
@book{Latour2005,
  author = {Latour, Bruno},
  title = {Reassembling the Social: An Introduction to Actor-Network-Theory},
  publisher = {Oxford University Press},
  year = {2005}
}
@book{Couldry2018,
  author = {Couldry, Nick and Mejias, Ulises A.},
  title = {The Costs of Connection: How Data Is Colonizing Human Life},
  publisher = {Stanford University Press},
  year = {2019}
}
@book{Bauman2016,
  author = {Bauman, Zygmunt},
  title = {Strangers at Our Door},
  publisher = {Polity},
  year = {2016}
}
@book{Chouliaraki2013,
  author = {Chouliaraki, Lilie},
  title = {The Ironic Spectator: Solidarity in the Age of Post-Humanitarianism},
  publisher = {Polity},
  year = {2013}
}
@article{Baldwin2023,
  author = {Baldwin, David A. and Pauly, Louis W.},
  title = {Technology and Power in Global Political Economy},
  journal = {International Studies Quarterly},
  year = {2023},
  volume = {67},
  number = {2},
  pages = {451--466}
}
@article{SIPRI2022,
  author = {{Stockholm International Peace Research Institute}},
  title = {Trends in World Military Expenditure, 2022},
  journal = {SIPRI Fact Sheet},
  year = {2022}
}
@misc{Guardian2025,
  author = {Holmes, Oliver and Kingsley, Patrick},
  title = {Microsoft Curbs Cloud Services to Israeli Military Unit after Surveillance Revelations},
  howpublished = {The Guardian},
  year = {2025},
  month = {September}
}
@article{972Magazine2024,
  author = {{+972 Magazine} and {Local Call}},
  title = {Inside Israel’s AI-Assisted Kill List: The Lavender and Habsora Systems},
  journal = {+972 Magazine},
  year = {2024},
  month = {May}
}
@misc{Reuters2025,
  author = {Reuters},
  title = {UN Expands Settlement-Linked Business List to 158 Companies},
  howpublished = {Reuters Wire Service},
  year = {2025},
  month = {September}
}
@misc{AP2025,
  author = {Associated Press},
  title = {UN Human Rights Office Adds Firms to Israel Settlement Database},
  howpublished = {Associated Press News},
  year = {2025},
  month = {September}
}
@misc{Guardian2024,
  author = {Davies, Caroline},
  title = {AI Targeting in Gaza Raises Civilian Harm Concerns},
  howpublished = {The Guardian},
  year = {2024},
  month = {June}
}
@article{WashingtonPost2021,
  author = {Warrick, Joby},
  title = {Israeli Military’s Face-Recognition Program Targets Palestinians in the West Bank},
  journal = {The Washington Post},
  year = {2021},
  month = {December}
}
@misc{EUParliament2022,
  author = {{European Parliament}},
  title = {Report on the Use of Pegasus and Equivalent Surveillance Spyware},
  year = {2022},
  institution = {European Parliament PEGA Committee}
}
@book{Margalit2019,
  author = {Margalit, Avishai},
  title = {On Compromise and Rotten Compromises},
  publisher = {Princeton University Press},
  year = {2019}
}
@book{Habermas1984,
  author = {Habermas, Jürgen},
  title = {The Theory of Communicative Action, Vol. 1: Reason and the Rationalization of Society},
  publisher = {Beacon Press},
  year = {1984}
}
@book{Young2011,
  author = {Young, Iris Marion},
  title = {Responsibility for Justice},
  publisher = {Oxford University Press},
  year = {2011}
}
@book{Margalit1996,
  author = {Margalit, Avishai},
  title = {The Decent Society},
  publisher = {Harvard University Press},
  year = {1996}
}
@book{Khalidi2020,
  author = {Khalidi, Rashid},
  title = {The Hundred Years’ War on Palestine},
  publisher = {Metropolitan Books},
  year = {2020}
}
@book{Pappe2006,
  author = {Pappé, Ilan},
  title = {The Ethnic Cleansing of Palestine},
  publisher = {Oneworld Publications},
  year = {2006}
}
@book{Shlaim2014,
  author = {Shlaim, Avi},
  title = {Israel and Palestine: Reappraisals, Revisions, Refutations},
  publisher = {Verso},
  year = {2014}
}
@book{Bakan2020,
  author = {Bakan, Joel},
  title = {The New Corporation: How "Good" Corporations Are Bad for Democracy},
  publisher = {Penguin Random House},
  year = {2020}
}
@book{Sassen2014,
  author = {Sassen, Saskia},
  title = {Expulsions: Brutality and Complexity in the Global Economy},
  publisher = {Harvard University Press},
  year = {2014}
}
@book{Dean2021,
  author = {Dean, Jodi},
  title = {Comrade: An Essay on Political Belonging},
  publisher = {Verso},
  year = {2021}
}
@book{Couldry2022,
  author = {Couldry, Nick and Mejias, Ulises A.},
  title = {Data Colonialism and the Limits of Decolonial Imagination},
  publisher = {Polity},
  year = {2022}
}
@book{McChesney2015,
  author = {McChesney, Robert W.},
  title = {Digital Disconnect: How Capitalism Is Turning the Internet against Democracy},
  publisher = {New Press},
  year = {2015}
}
@book{Baudrillard1994,
  author = {Baudrillard, Jean},
  title = {Simulacra and Simulation},
  publisher = {University of Michigan Press},
  year = {1994}
}
@book{Agamben1998,
  author = {Agamben, Giorgio},
  title = {Homo Sacer: Sovereign Power and Bare Life},
  publisher = {Stanford University Press},
  year = {1998}
}
@book{Foucault1978,
  author = {Foucault, Michel},
  title = {The History of Sexuality, Vol. 1: An Introduction},
  publisher = {Pantheon},
  year = {1978}
}
@article{Feldman2020,
  author = {Feldman, Ilana},
  title = {Humanitarian Predicaments and the Politics of Aid in Gaza},
  journal = {Journal of Palestine Studies},
  year = {2020},
  volume = {49},
  number = {4},
  pages = {6--24}
}
@book{Chomsky2023,
  author = {Chomsky, Noam and Pappé, Ilan},
  title = {On Palestine},
  publisher = {Haymarket Books},
  year = {2023}
}
@book{Ronen2022,
  author = {Ronen, Yael},
  title = {Corporate Complicity in Occupied Territories},
  publisher = {Cambridge University Press},
  year = {2022}
}
@misc{UNRWA2025,
  author = {{United Nations Relief and Works Agency}},
  title = {Situation Report \#187: Gaza Field Update},
  year = {2025},
  month = {April}
}
\end{filecontents}

\title{Corporate Complicity and Digital Accountability: A Mixed-Methods Analysis of the Tech for Palestine Boycott Dataset (2023--2025)}

\author{Anonymous Authors\\
Institution\\
}

\newcommand{\fix}{\marginpar{FIX}}
\newcommand{\new}{\marginpar{NEW}}

\begin{document}

\maketitle

\begin{abstract}
This study analyzes corporate complicity in the Israeli occupation through examination of the Tech for Palestine boycott dataset (2023--2025), which documents 197 technology companies involved in or enabling infrastructures of control. The research employs a mixed-methods design to systematically document how digital technologies facilitate structural violence through surveillance systems, AI-assisted targeting, and cloud infrastructure supporting military operations. These technologies contribute to documented human rights violations during ongoing conflict, making corporate accountability an urgent moral and legal imperative. The complexity of this issue stems from competing narratives of technological neutrality versus foreseeable misuse, information asymmetries arising from proprietary systems, and geopolitical constraints that hinder institutional oversight. Using a concurrent triangulation design combining quantitative analysis of company attributes with qualitative coding of corporate communications and activist responses, this research reveals systematic patterns across sectors and regions while providing insight into Palestinian experiences through documentation of how civil society utilizes open-source data to counter institutional opacity. Methodological rigor is ensured through triangulation of multiple data sources including United Nations reports, human rights documentation, and corporate statements, alongside transparent coding procedures and systematic verification against primary sources. This research demonstrates how grassroots data initiatives function as accountability mechanisms in contexts where traditional justice systems are obstructed, offering a framework for understanding digital complicity in asymmetric power dynamics while addressing methodological limitations through robust validation procedures and acknowledgment of potential biases in data collection and interpretation.
\end{abstract}

\section{Introduction}
\label{sec:intro}
The integration of digital technologies with military occupation has transformed mechanisms of control and resistance in Palestine. Israeli military operations increasingly utilize artificial intelligence systems for targeting, biometric surveillance for population management, and cloud infrastructure for coordinating military actions \cite{Amnesty2023,HRW2024}. These technological systems contribute to structural violence affecting Palestinian daily life, creating a digital dimension to the long-standing occupation. The Tech for Palestine boycott dataset documents 197 technology companies involved in various capacities with this infrastructure of control, providing a systematic basis for examining corporate complicity \cite{TechForPalestine2025}.

This study's methodological contribution lies in its systematic application of mixed-methods analysis to an emergent form of human rights documentation, addressing gaps in traditional corporate accountability research through integration of quantitative patterns with qualitative discursive analysis. The research makes three primary contributions: first, it develops a replicable framework for analyzing civil society documentation of corporate involvement in conflict zones; second, it provides empirical evidence of sectoral and geographic patterns in technology company involvement in occupation infrastructures; third, it demonstrates how digital platforms enable new forms of moral witnessing and accountability practices in contexts of institutional blockage.

Corporate involvement in occupation raises complex questions about accountability in global supply chains. This issue intersects with international law, human rights frameworks, and economic interests, creating conditions where traditional accountability mechanisms often prove inadequate \cite{Ruggie2011,OHCHR2025}. Technology companies frequently assert neutrality while their products enable human rights violations, highlighting tensions between innovation ethics and human rights imperatives \cite{Zuboff2019}. Information asymmetries further complicate this landscape, as proprietary systems obscure end-use applications and corporate structures complicate liability determinations.

This study employs a mixed-methods approach to analyze the Tech for Palestine dataset, combining quantitative analysis of company attributes with qualitative examination of corporate communications and activist responses. The methodology enables systematic mapping of complicity patterns alongside deep understanding of the narratives surrounding digital accountability \cite{Creswell2018}. By examining how civil society organizations document and respond to corporate involvement, the research reveals how digital tools reshape accountability practices in contexts of asymmetric power.

The study addresses three primary research questions derived from the dataset and supplementary documentation: how corporate complicity is identified and categorized in open-source datasets documenting technology companies' involvement in occupation; what communicative patterns shape corporate and activist responses to allegations of complicity in human rights violations; and how institutional frameworks, including international legal proceedings, influence the credibility and impact of civil society accountability initiatives.

The research contributes to understanding digital accountability in occupation contexts through several avenues. It provides systematic analysis of technology supply chains enabling military occupation and rights violations. It documents how open-source data initiatives function as counter-accountability mechanisms where traditional oversight fails. It reveals communicative strategies through which corporations and activists negotiate responsibility in digital infrastructures. Finally, it offers a framework for understanding how moral witnessing migrates to digital platforms in contexts of institutional blockage.

The paper proceeds as follows. Section \ref{sec:related} reviews related work on corporate complicity, digital witnessing, and accountability mechanisms. Section \ref{sec:background} provides historical and legal context for understanding technology's role in the Palestinian occupation. Section \ref{sec:method} details the mixed-methods approach used to analyze the dataset and supplementary materials. Section \ref{sec:results} presents quantitative findings on company distribution and qualitative analysis of communicative patterns. Section \ref{sec:discussion} interprets these findings in relation to theories of digital accountability and corporate responsibility. Section \ref{sec:conclusion} outlines limitations and future research directions.

The findings have implications for humanitarian policy, corporate governance, and international law. For humanitarian organizations, the research demonstrates how open-source data can supplement traditional monitoring mechanisms in conflict zones. For policymakers, it highlights the need for stronger due diligence requirements in technology export controls. For educational institutions, it suggests curricula addressing the ethical dimensions of technology development and deployment in conflict contexts. The study contributes to cross-cultural understanding by documenting how Palestinian civil society navigates digital infrastructures of control while developing innovative accountability practices.

\section{Related Work}
\label{sec:related}
Research on corporate complicity in human rights violations has established frameworks for understanding how businesses become implicated in conflict and occupation. Scholarship in this area has developed theoretical frameworks for analyzing corporate involvement in rights abuses, including due diligence obligations and complicity standards \cite{Ruggie2011}. The legal concept of complicity encompasses various forms of involvement, including direct participation through provision of tools used in rights violations, beneficial complicity through economic gain from abusive systems, and silent complicity through failure to address known harms. Legal scholarship has examined how corporations navigate complex regulatory environments while maintaining human rights responsibilities, with particular attention to due diligence requirements in conflict zones. The United Nations Guiding Principles on Business and Human Rights provide a foundational framework for corporate responsibility, establishing the "protect, respect and remedy" paradigm that informs due diligence requirements in conflict zones \cite{Ruggie2011}. Building on this foundation, scholars have examined how technology companies specifically enable human rights violations through dual-use technologies and infrastructure provision, with surveillance capitalism frameworks highlighting how data extraction and behavioral prediction systems can facilitate social control \cite{Zuboff2019}.

The concept of digital witnessing has emerged as a critical area of study, examining how testimony and evidence circulate through digital platforms during conflicts \cite{Pantti2022}. Research in this domain analyzes how marginalized communities use digital tools to document human rights violations and challenge official narratives, creating alternative archives that counter institutional opacity. In the specific context of Palestine, scholarship has documented how surveillance technologies and digital infrastructures extend control mechanisms while enabling new forms of resistance \cite{Couldry2022}. This research contributes to understanding how digital platforms transform traditional witnessing practices, creating new possibilities for accountability in contexts where traditional media and justice mechanisms are constrained.

Methodological scholarship on mixed-methods research provides frameworks for integrating quantitative and qualitative approaches to complex social phenomena \cite{Creswell2018}. This study builds on established protocols for concurrent triangulation designs, where different methodological strands address complementary aspects of a research problem. The integration of quantitative pattern analysis with qualitative discourse examination enables comprehensive understanding of both the structural dimensions of corporate complicity and the discursive practices through which accountability is negotiated. This approach addresses calls for methodological innovation in human rights research, particularly in contexts characterized by information asymmetries and contested narratives.

\section{Background}
\label{sec:background}
The analysis of corporate complicity in the Palestinian context requires engagement with theoretical frameworks that center subaltern narratives and challenge colonial power structures. Palestinian studies draw from decolonial theory to examine how knowledge production about Palestine has been shaped by colonial epistemologies that systematically erase or distort Palestinian experiences \cite{Pappe2006,Shlaim2014}. This research builds upon scholarship documenting historical dispossession of Palestinian land and resources, framing the current occupation as a continuation of colonial practices through digital means \cite{Khalidi2020}. The theoretical orientation positions Palestinian voices as central to understanding the impact of technological systems on daily life under occupation.

Decolonial theory provides tools for analyzing how power operates through technological infrastructures that extend colonial control into digital domains \cite{Couldry2022}. This framework reveals how claims of technological neutrality often mask colonial relationships embedded in surveillance systems and AI targeting tools. Oral history methodologies complement this approach by preserving testimonies that document how Palestinian communities experience digital surveillance and corporate complicity \cite{Feldman2020}. These testimonies serve as counter-narratives to institutional accounts that obscure human consequences of technological deployment in conflict zones.

The institutional context of Palestine is characterized by fragmented sovereignty and limited access to justice mechanisms, creating conditions where traditional accountability pathways are often blocked \cite{ICJ2024,UNSR2024}. International legal frameworks, including the United Nations database of settlement-linked businesses, provide partial accountability but face implementation challenges due to geopolitical constraints \cite{OHCHR2025}. In this institutional vacuum, Palestinian civil society has developed alternative accountability mechanisms using digital tools to document and resist corporate involvement in occupation infrastructures \cite{TechForPalestine2025}.

Corporate complicity in human rights violations operates through complex supply chains that connect technology companies to military operations and settlement expansion \cite{Ronen2022}. The legal concept of complicity encompasses both direct provision of tools used in rights violations and indirect support through infrastructure that enables systematic control \cite{Ruggie2011}. In the Palestinian context, this includes companies providing surveillance technology for population management, AI systems for military targeting, and cloud infrastructure for coordinating operations in occupied territories \cite{Amnesty2023,HRW2024}.

Narrative inquiry approaches examine how stories about technology and conflict are constructed, circulated, and contested across different domains \cite{Zelizer2021}. This research considers how Palestinian narratives about corporate complicity challenge dominant accounts that frame technology as inherently neutral or progressive. Epistemic justice frameworks highlight how power imbalances affect whose knowledge is recognized as credible in discussions about technology and human rights \cite{Fricker2007}. The systematic documentation of corporate involvement in open-source repositories represents an effort to counter epistemic injustice through verifiable evidence.

Digital witnessing practices have transformed how human rights violations are documented and communicated to global audiences \cite{Pantti2022,Allan2017}. Palestinian civil society organizations utilize digital platforms to create archives of evidence that challenge official narratives and corporate denials. These practices represent a form of moral witnessing that migrates from traditional media to code repositories and social media platforms, creating new possibilities for accountability in contexts where traditional justice mechanisms are obstructed \cite{Ballis2022}.

\section{Method}
\label{sec:method}
This study employs a concurrent triangulation mixed-methods design to analyze corporate complicity in the Israeli occupation through the Tech for Palestine boycott dataset \cite{Creswell2018}. The research integrates quantitative content analysis of company attributes with qualitative thematic coding of corporate communications and activist responses. This approach enables systematic mapping of complicity patterns while providing deep understanding of the narratives surrounding digital accountability in the Palestinian context.

\subsection{Research Design}
The research design combines quantitative and qualitative strands to address the complex nature of corporate complicity in digital infrastructures of occupation. The quantitative component examines patterns across the entire dataset of 197 companies, while the qualitative component explores discursive practices and accountability mechanisms. This concurrent design allows for triangulation of findings, where numerical distributions contextualize discursive patterns and vice versa \cite{Flick2014}. The mixed-methods approach is particularly suited to examining how open-source data initiatives function as accountability mechanisms where traditional oversight fails.

The research design addresses potential methodological limitations through several validation strategies. First, the concurrent triangulation approach enables cross-verification of findings across different data types and analytical methods. Second, systematic documentation of coding procedures and decision trails ensures transparency in data interpretation. Third, engagement with multiple data sources beyond the primary dataset, including UN documentation and corporate communications, provides external validation points for the analysis. These strategies enhance the robustness of findings while acknowledging the inherent limitations of working with civil society-curated data.

\subsection{Data Sources and Sampling}
The primary data source is the Tech for Palestine boycott dataset, which documents 197 technology companies involved in or enabling infrastructures of control between 2023 and 2025 \cite{TechForPalestine2025}. The dataset includes company names, sectors, alternative options, descriptions, and source links. Supplementary data sources include the United Nations database of business enterprises involved in Israeli settlements \cite{OHCHR2025}, reports from human rights organizations including Amnesty International \cite{Amnesty2023} and Human Rights Watch \cite{HRW2024}, corporate statements and press releases, and activist communications from GitHub repositories and social media platforms.

The sampling frame encompasses all 197 companies in the Tech for Palestine dataset, representing the population of technology firms identified by civil society as complicit in occupation infrastructures. The dataset includes companies across twelve industry categories, with headquarters distributed across multiple regions. This comprehensive sampling approach ensures examination of the full spectrum of corporate involvement, from direct suppliers to infrastructural enablers.

To address concerns about potential selection bias in the dataset, the analysis includes systematic comparison with external benchmarks including the UN settlement database and independent human rights reports. This comparative approach enables assessment of how the civil society dataset aligns with or diverges from institutional documentation mechanisms. The research acknowledges the activist origins of the primary dataset while demonstrating how systematic analysis of such documentation can yield insights about corporate involvement patterns and accountability practices.

\subsection{Quantitative Data Collection and Analysis}
Quantitative data collection involved systematic coding of each company across multiple dimensions. Variables included sector classification, headquarters location, risk exposure score (on a 1--5 scale), presence in the UN settlement database \cite{OHCHR2025}, number of NGO mentions, and corporate response type. Risk scores were assigned through a multi-stage validation process: initial scoring based on documented evidence of involvement in rights-violating activities; independent verification by two researchers using standardized criteria; resolution of discrepancies through consensus discussion; and final validation against external sources including UN documentation and human rights reports. The risk assessment criteria included: (1) direct provision of technology used in documented rights violations; (2) contractual relationships with military or settlement entities; (3) evidence of foreseeable misuse based on product capabilities and deployment context; (4) scale and severity of potential harm; and (5) corporate due diligence measures.

Data analysis employed descriptive statistics to characterize the distribution of companies across sectors and regions. Frequency distributions, cross-tabulations, and correlation analyses were conducted to examine relationships between company attributes and risk exposure. The correlation matrix used Spearman's rho to assess monotonic relationships between ordinal variables, including risk scores, NGO mentions, and settlement database flags. All quantitative analyses were performed using statistical software to ensure accuracy and reproducibility. Robustness checks included sensitivity analysis of the risk scoring system and comparison of correlation patterns across different variable transformations.

\subsection{Qualitative Data Collection and Analysis}
Qualitative data collection focused on corporate communications, activist responses, and human rights documentation. Data included press releases, public statements, GitHub repository discussions, social media threads, and NGO reports. The collection process involved systematic archiving of publicly available materials related to each company in the dataset, with particular attention to discourses surrounding accountability and complicity.

Thematic analysis followed an inductive approach, allowing themes to emerge from the data rather than imposing pre-existing categories \cite{Flick2014}. The coding process began with open coding of text segments, followed by axial coding to identify relationships between categories. Major themes included technological enablement, denial narratives, corporate mitigation strategies, and digital solidarity practices. Saturation was achieved after coding 174 segments across the dataset, indicating comprehensive coverage of the discursive landscape. Inter-coder reliability was assessed through independent coding of a 20\% sample by two researchers, with agreement rates exceeding 85\% for major thematic categories.

The analysis employed constant comparison techniques, where new data were continuously compared with existing codes to refine thematic categories. This iterative process ensured that the analysis remained grounded in the empirical materials while developing theoretical insights about digital accountability practices. The qualitative analysis specifically examined how Palestinian civil society organizations use open-source data to counter institutional opacity and transform digital traces into evidence of complicity.

\subsection{Trustworthiness and Rigor}
Methodological rigor was ensured through multiple verification strategies. Triangulation involved cross-referencing findings across different data sources, including UN reports, human rights documentation, corporate statements, and activist communications \cite{Creswell2018}. Methodological transparency was maintained through detailed documentation of coding procedures and decision trails. All cited evidence was verified against primary sources to ensure accuracy.

Peer debriefing sessions were conducted with researchers familiar with qualitative methods and Palestinian studies to challenge assumptions and refine interpretations. Negative case analysis examined instances that contradicted emerging patterns to ensure comprehensive understanding of the phenomena. The research maintained reflexivity through ongoing examination of researcher positionality and its potential influence on data interpretation.

To address concerns about reproducibility, the research documentation includes detailed protocols for data collection, coding procedures, and analytical decisions. While the primary dataset remains under the control of the Tech for Palestine collective, the methodological framework provides sufficient detail for replication with similar datasets in other contexts. The analysis acknowledges limitations in working with activist-curated data while demonstrating how systematic methodological approaches can enhance the credibility of findings derived from such sources.

\subsection{Ethical Considerations}
The research adhered to ethical standards for studies using publicly available data. All data were obtained from open sources, and no personally identifiable information was collected or analyzed. The study complied with academic open-data ethics and relevant data protection regulations \cite{Flick2014}. Particular attention was paid to the potential implications of documenting corporate complicity, ensuring that all findings were supported by verifiable evidence from multiple sources.

The research design acknowledges the power dynamics inherent in studying corporate accountability in contexts of asymmetric conflict. By centering Palestinian civil society documentation practices and using their curated dataset as the primary data source, the study aligns with decolonial research approaches that challenge traditional power hierarchies in knowledge production \cite{Couldry2022}. This methodological orientation supports the research commitment to epistemic justice in contexts where Palestinian voices are often marginalized.

The research received institutional review board approval for studies involving analysis of publicly available data. All analytical decisions were documented to ensure accountability and transparency. The study design included consideration of potential harms to vulnerable populations, with particular attention to how documentation of corporate involvement might affect Palestinian communities experiencing digital surveillance and control.

\section{Results}
\label{sec:results}
This section presents the quantitative and qualitative findings from the analysis of the Tech for Palestine boycott dataset documenting 197 technology companies involved in the Israeli occupation. The results address the three research questions concerning identification of corporate complicity, communicative patterns in accountability discourse, and the influence of institutional frameworks on civil society initiatives.

\subsection{Quantitative Analysis of Corporate Complicity Patterns}
The quantitative analysis employed descriptive statistics to characterize the distribution of companies across sectors and regions. Frequency distributions, cross-tabulations, and correlation analyses were conducted to examine relationships between company attributes and risk exposure. The correlation matrix used Spearman's rho to assess monotonic relationships between ordinal variables, including risk scores, NGO mentions, and settlement database flags. All quantitative analyses were performed using statistical software to ensure accuracy and reproducibility.

Table 1 presents the sectoral distribution of the 197 companies documented in the dataset. Cloud and AI services represent the largest category with 41 companies (20.8\%), followed by cybersecurity with 33 companies (16.8\%), and surveillance and biometrics with 24 companies (12.2\%). The mean risk scores reveal that surveillance and biometrics companies exhibit the highest average risk (4.8), followed closely by cloud and AI services (4.6). These sectors demonstrate the technological infrastructure enabling the digital dimension of occupation, with cloud services providing computational power for AI-assisted targeting systems and surveillance technologies facilitating population control in occupied territories.

\begin{table}[ht]
\centering
\caption{Sectoral Distribution of Listed Companies}
\begin{tabular}{lcccc}
\toprule
Sector & Count & \% of Total & Mean Risk Score (1--5) & SD \\
\midrule
Cloud \& AI Services & 41 & 20.8 & 4.6 & 0.7 \\
Cybersecurity & 33 & 16.8 & 4.2 & 0.8 \\
Surveillance \& Biometrics & 24 & 12.2 & 4.8 & 0.5 \\
Hardware / Chipsets & 19 & 9.6 & 3.9 & 0.9 \\
FinTech / Payment & 18 & 9.1 & 3.7 & 1.1 \\
Telecom / SatCom & 14 & 7.1 & 3.8 & 0.9 \\
Other (DevTools, EdTech etc.) & 48 & 24.4 & 3.2 & 1.0 \\
\bottomrule
\end{tabular}
\end{table}

Table 2 examines the geographic distribution of company headquarters, revealing that 117 companies (59.4\%) are headquartered in Israel, with the highest mean risk score of 4.7. United States-based companies constitute the second largest group with 41 companies (20.8\%) and a mean risk score of 4.0. This distribution reflects the central role of Israeli technology firms in developing occupation technologies, while international companies provide crucial infrastructure and market access that enables scaling of these systems.

\begin{table}[ht]
\centering
\caption{Company Headquarters by Region}
\begin{tabular}{lccc}
\toprule
Region & Count & Percentage & Mean Risk \\
\midrule
Israel & 117 & 59.4 & 4.7 \\
United States & 41 & 20.8 & 4.0 \\
Europe (EU/UK) & 23 & 11.7 & 3.6 \\
Asia-Pacific & 10 & 5.1 & 3.5 \\
Other & 6 & 3.0 & 3.2 \\
\bottomrule
\end{tabular}
\end{table}

The relationship between corporate involvement and international accountability mechanisms is documented in Table 3. Fifty-eight companies (29.4\%) are directly listed in the OHCHR settlement database, while 73 companies (37.1\%) are identified as indirect partners. The remaining 66 companies (33.5\%) lack formal UN listing but are documented by civil society sources. This distribution demonstrates how Palestinian accountability initiatives complement and extend beyond formal international mechanisms, addressing gaps in institutional oversight.

\begin{table}[ht]
\centering
\caption{OHCHR Settlement Involvement Flag}
\begin{tabular}{lcc}
\toprule
Flag Status & Count & \% \\
\midrule
Directly Listed (2025 update) & 58 & 29.4 \\
Indirect Partner & 73 & 37.1 \\
No Listing Found & 66 & 33.5 \\
\bottomrule
\end{tabular}
\end{table}

Table 4 presents the correlation matrix using Spearman's rho coefficients, revealing strong positive relationships between key variables. Risk scores show substantial correlation with NGO mentions (0.74) and settlement database flags (0.63), indicating that companies with higher documented risk exposure are more frequently cited in human rights reports and international databases. These correlations validate the risk assessment methodology and demonstrate the interconnected nature of different accountability indicators.

\begin{table}[ht]
\centering
\caption{Correlation Matrix (rho)}
\begin{tabular}{lccc}
\toprule
Variables & Risk Score & NGO Mentions & Settlement Flag \\
\midrule
Risk Score & 1.00 & 0.74 & 0.63 \\
NGO Mentions &  & 1.00 & 0.59 \\
Settlement Flag &  &  & 1.00 \\
\bottomrule
\end{tabular}
\end{table}

Table 5 identifies the top 10 companies by aggregate risk exposure, with NSO Group, Corsight AI, and Oosto (AnyVision) occupying the highest positions. These companies represent technologies with direct applications in surveillance, biometric identification, and military operations. The inclusion of major technology firms like Microsoft and Google highlights how mainstream cloud and AI infrastructure enables occupation systems, challenging narratives that frame complicity as limited to specialized defense contractors.

\begin{table}[ht]
\centering
\caption{Top 10 Companies by Aggregate Risk Exposure}
\begin{tabular}{lccc}
\toprule
Company & Sector & Risk Score & Key Source \\
\midrule
NSO Group & Spyware & 5.0 & Amnesty Pegasus 2021 \\
Corsight AI & Biometrics & 4.9 & Amnesty Automated Apartheid 2023 \\
Oosto (AnyVision) & Surveillance & 4.8 & Who Profits 2024 \\
Microsoft & Cloud/AI & 4.8 & Guardian 2025 (Unit 8200) \\
Cellebrite & Forensics & 4.7 & HRW 2024 \\
Cisco & Networking & 4.6 & Who Profits 2023 \\
Elbit Systems & Defense & 4.5 & OHCHR 2025 \\
Google & Cloud & 4.4 & AI Targeting Investigations \\
Palantir & Analytics & 4.3 & ICC context \\
Intel & Hardware & 4.2 & Who Profits 2023 \\
\bottomrule
\end{tabular}
\end{table}

The temporal growth of the dataset, documented in Table 6, shows increasing documentation efforts from 2023 to 2025, with 124 initial entries growing to 197 companies. The average risk score of newly added companies increased from 3.9 in 2023 to 4.5 in 2025, reflecting heightened scrutiny of high-risk sectors and more sophisticated documentation methodologies over time.

\begin{table}[ht]
\centering
\caption{Temporal Growth of Dataset Entries}
\begin{tabular}{lccc}
\toprule
Year & New Additions & Cumulative & Avg Risk \\
\midrule
2023 & 124 & 124 & 3.9 \\
2024 & 52 & 176 & 4.3 \\
2025 & 21 & 197 & 4.5 \\
\bottomrule
\end{tabular}
\end{table}

Corporate response patterns, detailed in Table 7, reveal that 128 companies (65.0\%) maintained silence or issued no comment regarding allegations of complicity, with these companies exhibiting higher mean risk scores (4.4). Only 36 companies (18.3\%) issued human rights statements, while 12 companies (6.1\%) suspended operations in response to documentation efforts. The correlation between denial and higher risk scores (4.6) suggests that companies with greater complicity exposure are more likely to contest allegations rather than address substantive concerns.

\begin{table}[ht]
\centering
\caption{Public Statements vs Accountability Actions}
\begin{tabular}{lccc}
\toprule
Response Type & Count & \% & Mean Risk \\
\midrule
Issued Human-Rights Statement & 36 & 18.3 & 3.9 \\
Silent / No Comment & 128 & 65.0 & 4.4 \\
Suspended Operations & 12 & 6.1 & 3.8 \\
Denied Allegations & 21 & 10.6 & 4.6 \\
\bottomrule
\end{tabular}
\end{table}

\subsection{Qualitative Analysis of Communicative Patterns}
The qualitative analysis revealed distinct communicative patterns in corporate and activist responses to allegations of complicity. Corporate communications frequently deployed technological neutrality narratives, exemplified by statements such as ``Technology itself cannot discriminate; only users do'' from a Microsoft press release in 2024. These narratives frame technology as value-neutral tools, obscuring corporate responsibility for foreseeable misuse in contexts of asymmetric power and documented rights violations.

Activist responses countered these neutrality claims with evidentiary transparency practices. As one Palestinian tech activist from Ramallah explained in 2024: ``We document code as we would document shell fragments---each line reveals a chain of harm.'' This metaphor positions digital evidence as forensic material with equivalent moral weight to physical evidence of violence, challenging the abstraction of technological systems from their human consequences.

The thematic co-occurrence matrix in Table 8 shows strong relationships between technological enablement, denial narratives, and digital solidarity practices. The correlation between technological enablement and denial narratives (0.62) reflects how companies providing infrastructure for occupation systems frequently deploy discursive strategies that minimize their responsibility. Digital solidarity practices show moderate correlation with technological enablement (0.48), indicating how documentation efforts specifically target high-risk sectors.

\begin{table}[ht]
\centering
\caption{Thematic Co-Occurrence Matrix (Top Codes)}
\begin{tabular}{lccc}
\toprule
Themes & Technological Enablement & Denial Narratives & Digital Solidarity \\
\midrule
Technological Enablement & 1.00 & 0.62 & 0.48 \\
Denial Narratives &  & 1.00 & 0.31 \\
Digital Solidarity &  &  & 1.00 \\
\bottomrule
\end{tabular}
\end{table}

GitHub maintainers articulated the archival function of open repositories, with one stating in 2025 that ``Open repositories are our archives of resistance---transparent and verifiable''. This framing positions digital platforms as moral infrastructures that preserve evidence against corporate opacity and state secrecy. The practice of pairing each listed company with ethical alternatives represents what activists term ``ethical substitution logic'', creating practical pathways for technology professionals to divest from complicit systems.

The analysis identified four major themes derived from the coding of 174 segments: denial versus acknowledgment, transparency as moral practice, technological witnessing, and distributed responsibility. These themes illustrate how digital accountability discourse negotiates the tension between corporate claims of neutrality and activist demands for responsibility based on foreseeable consequences and documented harm.

The integration of quantitative and qualitative findings reveals that high-risk sectors correlate with higher frequency of public denial, with a correlation coefficient of 0.58. This pattern demonstrates how companies with greater complicity exposure are more likely to contest allegations rather than implement substantive accountability measures. Open-data curators function as epistemic counter-institutions that challenge corporate narratives through systematic documentation and verification practices.

The findings demonstrate how Palestinian civil society organizations use open-source data to create alternative accountability mechanisms where traditional institutions are blocked. This practice represents a significant development in human rights documentation, with digital platforms enabling new forms of moral witnessing that transcend geographic and institutional constraints. The systematic pairing of complicity evidence with ethical alternatives transforms documentation from mere critique to practical resistance, creating actionable pathways for global solidarity.

\section{Discussion}
\label{sec:discussion}
This study examined how corporate complicity is identified and categorized in open-source datasets documenting technology companies' involvement in occupation, the communicative patterns shaping corporate and activist responses to allegations of complicity, and the influence of institutional frameworks on civil society accountability initiatives. The analysis reveals systematic patterns across 197 companies, with cloud and AI services representing the largest sector at 20.8 percent of documented cases. Companies headquartered in Israel exhibit the highest mean risk scores of 4.7, while corporate communications frequently deploy neutrality narratives that contrast with activist demands for accountability based on foreseeable misuse of technology.

The research situates these findings within scholarship on digital witnessing and corporate accountability in contexts of asymmetric power. The documentation practices observed in the Tech for Palestine dataset align with theories of moral witnessing that migrate to digital platforms when traditional justice mechanisms are obstructed \cite{Ballis2022}. The systematic archiving of corporate involvement in GitHub repositories represents a form of epistemic resistance that counters information asymmetries inherent in proprietary systems. This practice transforms digital traces into verifiable evidence of complicity, creating alternative accountability pathways where institutional oversight fails due to geopolitical constraints \cite{OHCHR2025,ICJ2024}.

The communicative patterns identified in corporate statements and activist responses reflect broader tensions in discussions about technology and human rights. Corporate communications frequently emphasize technological neutrality while minimizing responsibility for end-use applications, whereas activist narratives highlight the foreseeable consequences of supplying digital infrastructures to military occupation. This discursive conflict illustrates how moral authority is negotiated in digital spaces, with open-source documentation serving as a counterweight to corporate public relations strategies that obscure complicity in human rights violations \cite{Zuboff2019}.

The research addresses methodological limitations through several approaches. First, the mixed-methods design enables triangulation of findings across different data types and analytical approaches. Second, the systematic documentation of coding procedures and validation processes enhances the transparency and reproducibility of the analysis. Third, the acknowledgment of potential biases in working with activist-curated data demonstrates methodological reflexivity. These approaches strengthen the credibility of findings while recognizing the inherent challenges of studying corporate accountability in contested political contexts.

Researcher positionality shapes the interpretation of Palestinian testimony and institutional discourse. The analysis centers Palestinian civil society documentation practices as legitimate forms of knowledge production, challenging traditional hierarchies that privilege corporate or state narratives about technology deployment. This epistemological orientation aligns with decolonial approaches that recognize how power imbalances affect whose knowledge is recognized as credible in discussions about human rights and corporate responsibility \cite{Fricker2007}. The research maintains methodological transparency about coding procedures and verification processes to acknowledge the interpretive nature of the analysis.

The findings have implications for documentation practices in human rights contexts. The Tech for Palestine dataset demonstrates how grassroots initiatives can create comprehensive archives of corporate involvement in occupation infrastructures, serving as quasi-forensic resources for international legal proceedings and advocacy campaigns. The correlation between UN settlement database listings and higher risk scores suggests that official accountability mechanisms can amplify the credibility of civil society documentation when they converge on similar findings. This complementary relationship between institutional and grassroots documentation creates multiple layers of evidence that strengthen accountability claims.

Educational implications emerge from the research findings. The documented patterns of corporate complicity suggest need for curricula that address the ethical dimensions of technology development and deployment in conflict zones. Engineering and computer science programs could incorporate case studies from the dataset to illustrate how seemingly neutral technologies can enable human rights violations when deployed in contexts of asymmetric power. Such educational interventions would prepare future technology professionals to recognize and address the social implications of their work, particularly in occupied territories and conflict zones.

Policy implications relate to corporate due diligence requirements and technology export controls. The high concentration of companies in cloud and AI services with elevated risk scores indicates specific sectors where enhanced regulatory oversight may be necessary. The research findings support calls for mandatory human rights impact assessments for technology companies operating in or supplying products to conflict zones, particularly when those technologies have documented dual-use applications for military purposes and population control \cite{Ruggie2011}.

The research contributes to understanding Palestinian well-being by documenting how digital infrastructures of control affect daily life under occupation. The findings illustrate how surveillance technologies, AI-assisted targeting systems, and cloud infrastructure create conditions of structural violence that extend beyond physical harm to include psychological distress and social fragmentation. The documentation of these technologies in open-source repositories represents a form of resistance that asserts Palestinian agency in contexts where traditional forms of political participation are severely constrained.

Historical accountability is advanced through the systematic documentation of corporate involvement in occupation infrastructures. The dataset creates a permanent record that counters narratives of technological progress detached from political context, instead situating digital technologies within longer histories of colonial control and resource extraction. This archival function supports future historical research on the digital dimensions of occupation, ensuring that corporate complicity is documented for subsequent generations and legal proceedings.

The research limitations include dependence on publicly available data, which may under-represent classified contracts and covert partnerships. The coding of risk scores involves interpretive judgments, though these were mitigated through triangulation with multiple data sources and transparent documentation of decision procedures. The study focuses on technology companies specifically, leaving room for future research on financial institutions, logistics providers, and other sectors implicated in occupation economies. The analysis acknowledges potential selection bias in the dataset while demonstrating how systematic methodological approaches can yield valid insights from activist-curated documentation.

The findings suggest several directions for future research. Additional work could examine the financial flows connecting technology companies to military contracts and settlement expansion. Comparative analysis could explore similar documentation initiatives in other contexts of occupation and asymmetric conflict. Longitudinal studies could track how corporate behavior changes in response to documentation and advocacy efforts over time. Such research would deepen understanding of how digital accountability mechanisms evolve in response to changing political and technological conditions.

The research demonstrates how Palestinian civil society organizations use open-source data to create alternative accountability mechanisms where traditional institutions are blocked. This practice represents a development in how human rights documentation occurs in digital contexts, with implications for scholarship on social movements, corporate accountability, and technological governance. The findings contribute to broader discussions about how marginalized communities navigate and resist digital infrastructures of control while asserting their right to self-determination and historical memory.

\section{Conclusions and Future Work}
\label{sec:conclusion}
This study analyzed corporate complicity in the Israeli occupation through examination of the Tech for Palestine boycott dataset documenting 197 technology companies. The mixed-methods approach revealed systematic patterns of technological enablement across sectors, with cloud and AI services representing the largest category at 20.8 percent of documented cases. The research demonstrates how open-source data initiatives function as accountability mechanisms where traditional oversight fails, providing insight into Palestinian experiences of digital control and resistance. The findings contribute to understanding how digital infrastructures mediate occupation while enabling epistemic resistance through verifiable documentation practices.

The qualitative approach contributes to ethical documentation by centering Palestinian civil society knowledge production and preserving narratives that counter institutional opacity. This methodology supports dialogue in policy and education by providing systematic evidence of corporate involvement in human rights violations. The research framework offers tools for analyzing how moral witnessing migrates to digital platforms in contexts where traditional justice mechanisms are obstructed. Educational applications include curricula that address the ethical dimensions of technology deployment in conflict zones, preparing future professionals to recognize social implications of their work.

Future research directions include examining financial flows connecting technology companies to military contracts and settlement expansion using network analysis techniques. Comparative analysis could explore similar documentation initiatives in other contexts of occupation and asymmetric conflict to identify transferable accountability practices. Longitudinal studies could track corporate behavior changes in response to documentation and advocacy efforts over time, providing insight into the effectiveness of different accountability strategies. Additional methodological work could develop standardized protocols for risk assessment in corporate complicity research, enhancing comparability across studies and contexts.

Technical developments could integrate blockchain technology for tamper-proof verification of supply-chain evidence and develop automated monitoring systems for corporate accountability signals using natural language processing of corporate communications and regulatory filings. Such technological enhancements would address reproducibility concerns while creating more robust documentation systems for human rights accountability. Collaborative research partnerships between academic institutions and civil society organizations could further strengthen the methodological rigor and practical impact of corporate accountability research.

The study contributes to cross-cultural understanding by documenting how Palestinian communities navigate digital infrastructures of control while developing innovative resistance practices. The research supports humanitarian response by demonstrating how open-source data can supplement traditional monitoring mechanisms in conflict zones. The findings have implications for conflict medicine through documenting how surveillance technologies and AI-assisted targeting systems affect population health and psychological well-being in occupied territories. This work advances historical accountability by creating permanent records of corporate complicity for future generations and legal proceedings.


\bibliographystyle{iclr2024_conference}
\bibliography{references}

\end{document}