Research
Computational Political Science & Extremism Studies
Research Overview
My research program sits at the intersection of computational social science, political violence, and technology studies. I develop and apply cutting-edge computational methods—machine learning, natural language processing, and social network analysis—to understand how online communities translate into offline political action.
Core Questions:
- How do digital platforms facilitate political radicalization and mobilization?
- What role do emerging technologies play in extremist recruitment and organization?
- How can computational methods help us understand and counter political violence?
Citation Metrics
Google Scholar Profile: 1,093 total citations | Research Areas: collective behavior, online communities, extremism, conspiracy theories, video games
Impact Highlights:
- Featured in 8+ major media outlets
- 276K+ total article views
- Research cited in policy reports and academic literature
Featured Publications
Peer-Reviewed Articles
Kowert, R., Kilmer, E., and Newhouse, A. (2024). “Taking it to the Extreme: Prevalence and Nature of Extremist Sentiment in Games.” Frontiers in Psychology, 15:1410620.
This study addresses the gap in research knowledge around extremist sentiment in games by evaluating prevalence, location, nature, and impact through an online survey of 423 game players, revealing alarmingly high rates of extremist content exposure and normalization within gaming cultures.
- Funded by: Department of Homeland Security (DHS # EMW-2022-GR-00036)
- Key Finding: More than half of all game players report experiencing some form of hate, harassment or abuse within gaming spaces, with extremist rhetoric becoming culturally normalized
Newhouse, A. (2021). “The Threat is the Network: The Multi-Node Structure of Neo-Fascist Accelerationism.” CTC Sentinel, 14(5).
Since 2015, the Atomwaffen Division has received bulk academic and media attention in coverage of the neo-fascist accelerationist movement, but evidence reveals it was not the apex of a hierarchy but rather one node in a larger network of violent accelerationists built on membership fluidity, frequent communications, and a shared goal of social destruction.
- Innovation: Novel network-based framework for understanding extremist movements
- Policy Impact: Enforcement against individuals and groups is necessary but not sufficient—focus on specific groups may not tackle the root of the issue
- Citations: Extensively cited in terrorism studies and cited in recent Studies in Conflict & Terrorism article
McGuffie, K. and Newhouse, A. (2020). “The Radicalization Risks of GPT-3 and Advanced Neural Language Models.” arXiv preprint, 2009.06807.
We show GPT-3’s strength in generating text that accurately emulates interactive, informational, and influential content that could be utilized for radicalizing individuals into violent far-right extremist ideologies and behaviors. While OpenAI’s preventative measures are strong, the possibility of unregulated copycat technology represents significant risk for large-scale online radicalization and recruitment.
- Impact: 28 citations and featured in major tech policy discussions
- Media Coverage: Extensive coverage in The Register, Vice, and other tech outlets
- Policy Influence: Already in 2020, Kris McGuffie and Alex Newhouse highlighted the potential for abuse of generative language models by assessing GPT-3, revealing significant risk for large-scale online radicalization and recruitment
Other Major Publications
Newhouse, A. (2020). “Far-right activists on social media telegraphed violence weeks in advance of the attack on the US Capitol.” The Conversation.
Newhouse, A. (2020). “Parler is bringing together mainstream conservatives, anti-Semites and white supremacists as the social media platform attracts millions of Trump supporters.” The Conversation.
- Impact: 241,000 views for Parler analysis
Multiple GNET Research Reports (2021-2022) on accelerationist movements and extremist narratives, including analysis of the Great Replacement theory and Boogaloo movement.
Research Areas & Methods
Substantive Focus Areas
Political Violence & Extremism
- Neo-fascist accelerationism and network structures
- Online-offline radicalization pathways
- Gaming platforms as extremist recruitment venues
- AI and synthetic media threats
Technology & Society
- Digital platform governance and content moderation
- AI safety and misuse potential
- Social media data analysis and behavioral patterns
- Gaming culture and political socialization
Methodological Expertise
Computational Methods
- Machine Learning (PyTorch, Scikit-Learn, Hugging Face Transformers)
- Natural Language Processing and text-as-data approaches
- Social Network Analysis (igraph, statnet)
- Time-series analysis and causal inference
Featured Technical Work:
- DistilBERT for Political Text Classification - Automated detection of extremist content using transformer models (94% F1-score)
Data & Platforms
- Large-scale social media data collection and analysis
- Gaming platform research and mixed-methods approaches
- Survey research and experimental design
- Qualitative coding and content analysis
Current Research Projects
Dissertation Research
“Measuring Online-Offline Relationships in Political Violence”
Developing computational frameworks to understand how digital engagement translates into offline political action. Uses machine learning, natural language processing, social network analysis, and time-series causal inference to examine extremist community dynamics.
- Committee: Andrew Q. Philips (co-chair), Jennifer Fitzgerald (co-chair), Alexandra Siegel
- External Advisor: Brian Keegan (Information Science)
- Expected Completion: May 2027
COVID-19 Media Framing Study
“Partisan Visual Politics During a Pandemic” (with Andrew Q. Philips and Komal P. Kaur)
Investigating partisan differences in visual and textual framing of pandemic coverage across U.S. news outlets using computer vision and text analysis methods.
Gaming & Extremism Research
“Multi-Platform Radicalization in Digital Gaming”
Comprehensive study of recruitment and radicalization processes within digital gaming environments, including direct partnerships with major gaming platforms for harm mitigation strategies.
Policy & Financial Markets
“Political Messaging and Market Responses” (under review)
Analysis of how political communications affect financial market behavior and investor decision-making.
Research Impact & Media Coverage
Academic Recognition
- Research extensively cited in terrorism studies literature
- Work featured in CTC Sentinel, premier counterterrorism publication
- Publications in top-tier psychology and computer science venues
Policy & Industry Impact
- January 6th Committee: Investigative consultant providing expertise on extremist movements
- Technology Partnerships: Direct collaboration with gaming companies (Roblox, Spectrum Labs)
- Federal Funding: $1.38M in grants as Principal Investigator
Media & Public Engagement
Major Media Features: Washington Post • New York Times • NPR (Morning Edition, All Things Considered) • BBC • Politico • Wired • Bloomberg Radio • Meet the Press Now
High-Impact Articles:
- Parler analysis: 241,000 views
- January 6th prediction: 35,400 views
Expert Commentary Topics:
- AI safety and misuse potential
- Gaming platform moderation
- Extremist recruitment strategies
- Social media radicalization
- Technology policy implications
Funding & Grants
As Principal Investigator
- Department of Homeland Security (2022-2024): Gaming and extremism research - $350,000
- Multiple Federal Grants (2019-2024): Extremism and technology studies - $1.38M total
As Co-Investigator
- Logically Partnership (2022): Social media data infrastructure for extremism research
- Various Industry Partnerships: Gaming platform safety research
Future Research Directions
Emerging Technologies & Political Violence
- AI-generated disinformation and radicalization
- Virtual/augmented reality environments as political spaces
- Blockchain and decentralized platform governance
Comparative Extremism Studies
- Cross-national analysis of digital radicalization patterns
- Comparative platform governance approaches
- International cooperation in countering online extremism
Methodological Innovation
- Advanced causal inference methods for social media data
- Multi-modal analysis combining text, visual, and network data
- Real-time detection and intervention systems
For collaboration inquiries, media requests, or access to datasets, contact alex.newhouse@colorado.edu.
Google Scholar: Alex Newhouse • ORCID: Available upon request