← Back to heesoojang.com

Research portfolio

Heesoo Jang

Assistant Professor of Media Law and Ethics · University of Massachusetts Amherst

My research asks who governs what a democracy knows. Journalism, platforms, and AI increasingly operate as interconnected governors of democratic knowledge, shaping what becomes publicly knowable and credible, whose expertise counts, and who bears responsibility when information systems fail.

Across the United States, South Korea, and other national contexts, I study how these institutions produce knowledge, authorize stakeholder voice, and exercise corporate and infrastructural power.

An editorial illustration connecting a ballot box, newspaper, courthouse, public conversation, and an abstract AI network.
Democratic knowledge is shaped across journalism, law, elections, platforms, and AI.
Published work / dissertation Conference paper

Selected work

Three places to begin

Three projects that introduce the program's central concerns: consequential omission, democratic accountability, and infrastructural dependency.

Public knowledgeLLMs and U.S. ElectionsAuditing not only what generative AI gets wrong, but what it leaves democratically unusable. JournalismDemocracy-Framed JournalismShowing how routine election coverage can normalize claims that reject democratic procedure. InfrastructureJournalism's Dependency TrapTracing how news institutions come to rely on AI infrastructure they cannot control, audit, or readily leave.

Cross-cutting expertise

Auditing opaque systems and measuring consequential absence

I design empirical studies of systems that cannot be directly observed and translate normative theory into measurable baselines. This makes it possible to study not only what institutional texts and technical outputs contain, but also which context, obligations, and stakeholder voices should have appeared and did not.

Computational text analysisMixed-method AI auditingSurveysInterviews and focus groupsQualitative and thematic analysisNormative and legal inquiry
Strand 1

AI, Journalism & Public KnowledgeFlagship

How AI and platform systems reshape journalism's economic, ethical, and infrastructural conditions, and what those changes do to public knowledge.

LLMs and U.S. ElectionsHow accurately and democratically do large language models represent U.S. elections?PublishedConference paper›
Peer-reviewed publication
  • Jang, H.*, McGregor, S. C.*, & Neill, L.* (2026). Inaccuracies, omissions, and bullshit: 2024 U.S. elections through the eyes of ChatGPT. Information, Communication & Society [Special issue]. Article
Conference paper
  • Jang, H.*, McGregor, S. C.*, & Neill, L.* (2025, September). Investigating the Democratic Alignment of LLMs during the 2024 U.S. Elections. APSA, Vancouver.

Contribution. Extends communicative gatekeeping theory to generative systems and develops democratic information standards for evaluating consequential omissions.

Democracy-Framed Journalism & the Legitimation of Far-Right NewsHow can established news routines normalize actors and claims that undermine democratic institutions?Published›
Peer-reviewed publications
  • Jang, H.*, & Kreiss, D.* (2025). Safeguarding the peaceful transfer of power: Pro-democracy electoral frames and journalist coverage of election deniers during the 2022 U.S. midterm elections. The International Journal of Press/Politics, 30(3), 775–796. Article Supplement
  • Archer, A. M. N., Schmitt, C., McGregor, S. C., & Jang, H. (2025). Presidential authority and the legitimation of far-right news. The International Journal of Press/Politics, 30(1), 14–37. Article Supplement

Contribution. Introduces democracy-framed electoral coverage and identifies presidential mentions as authority signals that can transfer institutional legitimacy to far-right media.

AI Infrastructure Dependency in JournalismWhat happens when journalism relies on technical infrastructure it does not control, cannot fully audit, and cannot easily exit?Conference paper›
Conference paper
  • Jang, H. (2026, April). Lessons from the platform era for AI governance: The infrastructural dependency trap in local journalism. Local Journalism Researchers Workshop, Duke in DC, Washington.
Amplification, Repair & Celebrity Journalism EthicsWhy can a correction fail to repair reputational harm after allegations have circulated across an attention economy?Conference paperAward-winning paper›
Research presentation
  • Jang, H. (2025, August). The Case of G-Dragon and the Ethics of Celebrity Journalism. AEJMC, San Francisco. First Place Faculty Paper
Strand 2

AI Governance, Law & Stakeholder Voice

Who is authorized to define AI in legal and policy settings, who is recognized as an expert, and whose experience counts as evidence.

A Theory of Stakeholder Voices for Legal AIBefore a technology is adopted, who gets to define what it is for and which consequences matter?Conference paperTop paper›
Conference papers
  • Jang, H.*, Reid, A.*, & Ringel, E.* (2025, August). Who gets to shape the future? A theory of stakeholder voices for understanding legal AI. Political Communication Division, AEJMC Annual Conference, San Francisco.
  • Jang, H.*, Reid, A.*, & Ringel, E.* (2025, March). Beyond innovation diffusion theory: Stakeholder perspectives in news coverage of legal AI adoption. Law & Policy Division, AEJMC Southeast Colloquium, Chapel Hill. Top Faculty Paper
Whose Voice Counts in News Coverage of Legal AI?How does news coverage authorize institutional experts while rendering people subject to AI-assisted legal decisions largely absent?Conference paper›
Conference paper
  • Jang, H., Ringel, E., & Reid, A. (2025, June). The future of justice: A computational mixed-method approach to news coverage of AI in the United States legal system. Law & Policy Division, ICA Conference, Denver.
First Amendment Theory in an Age of Synthetic MediaWhat happens to the marketplace-of-ideas metaphor when speech has no identifiable human speaker and expressive supply is effectively unbounded?Published›
Published law review
  • Schroeder, J., & Jang, H. (2026). A synthetic marketplace: Rethinking First Amendment theory in the age of AI-generated video. Business, Entrepreneurship & Tax Law Review, University of Missouri School of Law. Publication

Contribution. Argues that courts need an account of expressive value that does not depend entirely on attribution.

Strand 3

Corporate & Infrastructural Power

How firms and states convert the language of responsibility and autonomy into legitimacy, and how harms and obligations move across privately controlled infrastructures.

Layered Affordances & Platform-Enabled HarmHow do harms compound as people and content move across platforms with different features and moderation regimes?Published›
Peer-reviewed publication
  • Jang, H., & Narayanamoorthy, N. (2025). Echo chambers of digital harm: Insights into layered affordances from India and South Korea. Social Media + Society, 11(4), 20563051251383528. Article Supplement
Related work
  • Jang, H. (2023, October). Using Data Feminism to Study South Korea's Mass Digital Sex Trafficking Case. A Toolbox of Feminist Wonder Workshop, CSCW '23, Minneapolis.
  • Jang, H., et al. (2023). Platform (In)Justice: A Call for a Global Research Agenda. CSCW '23, Minneapolis.
  • Abidin, T., & Jang, H. (2026, April). Her face, their fantasy: How deepfakes exploit accessibility. UMass Amherst SBS Undergraduate Research Symposium, Amherst. Undergraduate mentee first author

Contribution. Shows why platform-by-platform content policy is insufficient for harms produced across X, Instagram, and Telegram, and why platform theory requires comparative, non-Western cases.

Public Expectations of AI Companies & Participatory GovernanceHow do knowledge, optimism, and populist attitudes shape demands for corporate obligation and public participation?Conference papers›
Three-country survey research
  • Xu, H., Zhu, Y. E., Jang, H., & Juarez Miro, C. (2026, August). Value-based communication in the AI industry: Public evaluation of AI companies' organization–public relationships and governance expectations. Public Relations Division, AEJMC Annual Conference, New Orleans.
  • Xu, H., Zhu, Y. E., Jang, H., & Juarez Miro, C. (2026, June). From optimism to obligation: How positive orientation toward AI drives public demand for corporate digital responsibility. Public Relations Division, ICA Annual Conference, Cape Town.
  • Zhu, Y. E., Juarez Miro, C., Jang, H., & Xu, H. (2026, June). Confidence engages; understanding withdraws: How knowledge and populist attitudes shape orientations toward AI and participatory governance across three countries. Political Communication Division, ICA Annual Conference, Cape Town.
Sovereign AI, Symbolic Autonomy & Material DependencyHow do governments narrate national control while depending on a small number of firms, supply chains, and geopolitical alliances?Conference paper›
Conference paper
  • Jang, H. (2025, October). Exporting autonomy, importing dependency: The geopolitical work of "sovereign AI." AAAI/ACM Conference on AI, Ethics, and Society, Madrid. [Non-archival track]
AI Ethics Discourse & Corporate ResponsibilityWhen institutions invoke ethics and responsibility, which harms and obligations become visible, and which remain outside enforceable accountability?DissertationConference paper›
Foundational work
  • Jang, H. (2024). An Integrative Framing Study of the Public Discourse around Artificial Intelligence (AI) Ethics [Doctoral dissertation, University of North Carolina at Chapel Hill]. Dissertation
Conference papers
  • Jang, H. (2025, August). The ethics we read: How news media shape the moral imagination of AI. PolNet-PaCSS, Boston/Cambridge.
  • Jang, H., & Cho, J. (2024, June). An assessment of reported biases and harms of large language models. Human-Machine Communication Interest Group, ICA Annual Conference, Gold Coast. Top Paper
  • Jang, H. (2023, November). Bringing AI under critical CSR scrutiny: A critical CSR assessment of major AI companies' communication of responsible AI. Public Relations Division, NCA Annual Conference, National Harbor.
  • Jang, H. (2023, March). What should CSR in the AI industry look like? A current assessment and a framework for the future. IPRRC, Orlando.

Finding. The language of ethics and responsibility expanded much faster than the scope of enforceable obligation.