About

Mike Ananny is an associate professor of communication and journalism at USC Annenberg. He is an interdisciplinary scholar working at the intersections of communication, journalism studies, media studies, and science and technology studies. His research traces how technologies and cultures of media production have the power to shape public life, with a focus on the practices, assumptions, and controversies driving social media platforms, data infrastructures, algorithms, and artificial intelligence. Ananny studies the future of public life through these sociotechnical systems and is involved in the faculty advisory committee of USC’s Center on Science, Technology, and Public Life. He co-leads the interdisciplinary research collective 'MASTS' (Media as SocioTechnical Systems) and has held numerous fellowships and scholarships with institutions such as Stanford University, Columbia University, Harvard University, and others. His academic background includes a PhD from Stanford University in communication, an SM from the MIT Media Lab in media arts and sciences, and a BSc from the University of Toronto in human biology and computer science. Ananny has authored and edited books on networked press freedom and Bauhaus futures, and regularly publishes in both academic and popular venues.

Research topics

• Computer Science
• Political Science
• Sociology
• Information Retrieval
• Data Mining
• Social psychology
• Media studies
• Data science
• Epistemology
• World Wide Web
• Public relations
• Knowledge management
• Psychology

Selected publications

• Data disaffection: Toward a relational and affective understanding of datafication

  New Media & Society · 2025-01-12 · 6 citations

  articleSenior author

  Research on user experiences with datafication, the transformation of social life into data, identifies “digital resignation” and “privacy cynicism” as rational responses to feeling overwhelmed and disempowered. But how, exactly, do shared feelings and emotions mediate relationships between datafication and disengaged responses – both individually and institutionally? We develop a relational analysis of datafication, deploying an infrastructural perspective and drawing on affect theory to develop the concept of data disaffection , which we define as the structural cultivation of accepting data accumulation as inevitable. Data disaffection is a structure of feeling that conditions processes across scales of analysis: it manifests in resignation and cynicism on an individual level while simultaneously structuring commercial practices. We illustrate how data disaffection highlights alternative sites and methods for understanding datafication, and we conclude by discussing the implications for understanding datafication as a cultural dynamic as well as a corporate practice.

  PublisherDOI

• AI Governance as scale work: synthetic journalism across scalar collisions

  Information Communication & Society · 2025-12-05 · 1 citations

  article1st authorCorresponding
  PublisherDOI

• HOW MEDIA UNIONS STABILIZE TECHNOLOGICAL HYPE Tracing Organized Journalism’s Discursive Constructions of Generative Artificial Intelligence

  Digital Journalism · 2025-01-25 · 9 citations

  article1st authorCorresponding
  PublisherDOI

• A Critical Field Guide for Working with Machine Learning Datasets

  ArXiv.org · 2025-01-26 · 5 citations

  preprintOpen access

  Machine learning datasets are powerful but unwieldy. Despite the fact that large datasets commonly contain problematic material--whether from a technical, legal, or ethical perspective--datasets are valuable resources when handled carefully and critically. A Critical Field Guide for Working with Machine Learning Datasets suggests practical guidance for conscientious dataset stewardship. It offers questions, suggestions, strategies, and resources for working with existing machine learning datasets at every phase of their lifecycle. It combines critical AI theories and applied data science concepts, explained in accessible language. Equipped with this understanding, students, journalists, artists, researchers, and developers can be more capable of avoiding the problems unique to datasets. They can also construct more reliable, robust solutions, or even explore new ways of thinking with machine learning datasets that are more critical and conscientious.

  PublisherOA PDFDOI

• How Media Unions Stabilize Technological Hype: Tracing organized journalism’s discursive constructions of Generative Artificial Intelligence

  2025-01-15

  preprintOpen access1st authorCorresponding

  Amidst ongoing challenges to journalism’s economic models, labor markets, and technological practices, a new pressure has recently appeared in many newsrooms: the power of Generative Artificial Intelligence (GenAI) computational models and off-the-shelf interfaces to synthetically create content that passes for news. Seeing the phenomenon through the lens of this special issue’s focus on “hype,” this paper uses discourse analysis to understand how journalism unions define GenAI as a problem, articulate the value of journalism against it, and use collective bargaining to contractually shape its use in newsrooms. Motivated by scholarship detailing hype as popular communication, expectation setting, and technological stabilizing, we examine journalistic trade press, union statements, and collective bargaining agreements to offer a 6-dimensional image of GenAI hype and union-driven responses to it, and reflect on notable absences in media unions’ understanding of GenAI. We see this as a case of journalists articulating their roles and values in an all-too-common moment when they are challenged by sociotechnical forces that they did not create, but that they must nonetheless collectively navigate and reshape in service of the profession’s democratic mission.

  PublisherOA PDFDOI

• Making Generative Artificial Intelligence a Public Problem. Seeing Publics and Sociotechnical Problem-Making in Three Scenes of AI Failure

  Javnost - The Public · 2024-01-02 · 14 citations

  article1st authorCorresponding

  Although Generative Artificial Intelligence (GenAI) has rapidly gained popular attention as something to worry about or be excited by, it is less clear if or how it is a public problem, and what ideals of publicness it illustrates, challenges, or invents. By developing the idea that GenAI is simultaneously an object and agent of public life, by describing how its failures are constructed and animated in three sociotechnical scenes, and by examining those scenes of failure for evidence of publicness, I trace how GenAI might be made into a public problem, and a problem for different ideals of publicness. Tracing how GenAI failures are narrated by charismatic figures, indexed by activists and policymakers, and avoided and repaired by journalists, I suggest that GenAI's public significance stems from its dual identity as both an ontological and epistemological concern and show how that duality plays out in failures that illustrate, combine, and extend ideals of the public.

  PublisherDOI

• Making events: How anticipatory infrastructures produce shared temporalities

  New Media & Society · 2024-03-15 · 1 citations

  articleSenior author

  Anticipatory infrastructures assemble sensors that are ready to detect, networks primed to share data, scientists prepared to confirm events, and news organizations poised to tell stories. This article explains how public time is articulated through sensor-mediated communications by examining two anticipatory infrastructures. Each infrastructure uses similar earthquake data to detect, report on, and convene material publics around earthquakes in Southern California. They are integral to structuring rhythms, coordinating syncronizations, setting deadlines, and making events timely, meaningful, and actionable, yet their governance lives in no one place. Instead, they emerge from an assemblage of sensors, networks, devices, algorithms, people, data, organizations, professional practices, and normative theories of the public. By comparing two different anticipatory infrastructures, we show how imagined publics, forms of journalistic storytelling, representations of earthquake events, and system maintenance can convene different public temporalities. We identify four dynamics involved in making these variable temporalities in material publics: how human-machine relations organize time, how professional norms of timeliness collide, how publics are anticipated by infrastructures, and how sensor infrastructures are maintained or decay over time.

  PublisherDOI

• Making Mistakes: Constructing Algorithmic Errors to Understand Sociotechnical Power

  2023-04-13 · 1 citations

  preprintOpen access1st authorCorresponding

  From law and politics to commerce and art, algorithms are powerful sociotechnical forces. But what does it mean when algorithms “fail”? What do we learn about sociotechnical dynamics when algorithms are seen to have erred or made a mistake? Seeing algorithms as culture, I argue that algorithmic errors are constructs of inter- twined computational, psychological, organizational, infrastructural, discursive, and normative forces. Through three stories of error, I show algorithmic failures as illustrations not only of algorithmic power but also of normative forces that de- fine success, rationalize iteration, and distribute harm. Instead of seeing algorithmic errors as unavoidable parts of technological innovation or self-evident transgressions, I instead see them as evidence of how people think systems should work, and the power to declare failures, trigger fixes, and envision futures by discovering and repairing mistakes. This power to “make mistakes" is a crucial and largely understudied form of sociotechnical control.

  PublisherOA PDFDOI

• An AI Society

  Issues in Science and Technology · 2023-12-28 · 3 citations

  articleSenior author

  Artificial intelligence is reshaping society, but human forces shape AI. Social scientists and humanities experts explore how to harness the interaction, revealing urgent avenues for research and policy.

  PublisherDOI

• An Infrastructural Approach to Digital Authoritarianism

  2023-05-22

  preprintOpen access1st authorCorresponding

  Starting from a standard working definition of digital authoritarianism (DA)---“the use of information technology by authoritarian regimes to surveil, repress, and manipulate domestic and foreign populations”---I want to suggest that the concept can more richly be seen as the unchecked creation and deployment of oppressive communication infrastructure. A focus on infrastructure offers a broader and more powerful way forward for counter-authoritarianism than the emphasis on tools and technologies that often dominates DA discussions. I try to illustrate this claim with two stories, a definition of DA infrastructure, and a short tour of places where DA infrastructure appears today. I end by suggesting three ways that an infrastructural view of digital authoritarianism could drive new approaches to counter-authoritarianism.

  PublisherOA PDFDOI

Frequent coauthors

• Kate Crawford

  Microsoft (United States)

  11 shared

• Carol Strohecker

  University of Minnesota

  7 shared

• Leila Bighash

  University of Arizona

  4 shared

• Kathleen Biddick

  Pennsylvania State University

  3 shared

• Mary L. Gray

  Microsoft (United States)

  3 shared

• Megan Finn

  James Cook University Hospital

  3 shared

• Frances Corry

  University of Pittsburgh

  3 shared

• Hamsini Sridharan

  University of Southern California

  3 shared

Awards & honors

• Faculty fellow with USC’s Society of Fellows in the Humaniti…
• Berggruen fellow at Stanford University’s Center for Advance…
• Fellow at Columbia University’s Tow Center for Digital Journ…
• Fellow and faculty associate at Harvard University’s Berkman…
• Doctoral scholar with Pierre Elliott Trudeau Foundation (200…