Skip to main content

Dr Andrew Dwyer

Addison Wheeler Research Fellow

AffiliationRoom numberTelephone
Addison Wheeler Research Fellow in the Department of Geography  
Addison Wheeler Research Fellow , Politics-State-Space  


Andrew's interests span a range of critical research on computational securities (cybersecurity), which incorporate malicious software, computational agency, and offensive cyber operations.

His current research considers how machine learning algorithms and humans arrive at decisions with regards to offensive (and defensive) cyber, and the implications of this for claims of 'automated' security. As part of this, he is Co-Lead of the newly formed academic Offensive Cyber Working Group, bringing together a diverse group of academics from across the UK and beyond to discuss and research offensive cyber. In further exploring critical perspectives to data, ethics, and computation, he is Co-I of a Human Data Interaction (HDI) Network-funded project, Zoom Obscura.

This research develops on his PhD - Malware Ecologies: A Politics of Cybersecurity – completed at the University of Oxford in 2019, which offers novel ways to think of malware through an autoethnographic study of the analysis laboratory of the endpoint detection vendor, Sophos. In this, he rethinks the relationship between environments, humans, and computation, using the work of N. Katherine Hayles, to argue that computation is not just a tool but rather a political actor in the negotiation and performance of cybersecurity.

Andrew is also a Research Affiliate at the Centre for Technology and Global Affairs at the University of Oxford, Assistant Editor at the journal Big Data & Society, and is on the Editorial Board for the journal, Digital Geography and Society. Previously he has been a Research Associate at the University of Bristol's Cyber Security Group and a Visiting Fellow at the SFB-TRR 138 'Dynamics of Security' collaborative research centre in central Germany.

Current Research

Andrew’s ongoing research fellowship – Digital Decisions (DIGIDEC) – seeks to question how the application of machine learning to cybersecurity – ‘Cyber AI’ – is revolutionising security practice through challenging the speed and efficacy of human decision. This project considers differing forms of (re)cognition, using N. Katherine Hayles, through combining insights from Geography, Science and Technology Studies (STS), International Relations, and beyond to offer a distinct challenge to the ‘automation’ of security. Security cannot be ‘hard-coded’ and is rather a socially negotiated process. What are then machine learning algorithms processing, and does it align with our contested expectations of security? As cybersecurity embraces machine learning across defensive and offensive postures, what are the dangers of such an approach? Computation exposes us to a world of securities that are neither reliant wholly on the human nor computation for their production, but a mixture between the two. The project then seeks to investigate the implications of differing forms of (re)cognition and ‘digital decisions’ for offensive cyber operations, adversarial machine learning, and future practices of war and security through detailed empirical study.

Media Contacts

Available for media contact about:

  • International: Defence & disputes:
  • Security, territory and boundaries:


Book review

Chapter in book

Doctoral Thesis

  • Dwyer, Andrew Carl (2019). Malware Ecologies: A Politics of Cybersecurity. University of Oxford. PhD.

Journal Article

Newspaper/Magazine Article

  • Dwyer, A. C. (2021). Why ‘Cyber Pearl Harbor’ matters for democracy. The Alert
  • Dwyer, A. C. (2021). Is the National Cyber Force for Good? Fabians' Defence & Security
  • Dwyer, Andrew C (2018). The NHS cyber-attack: A look at the complex environmental conditions of WannaCry. RAD Magazine (44): 25 - 26.

Other (Digital/Visual Media)

  • Dwyer, Andrew C (2020). Algorithms Don't Make Decisions!
  • Dwyer, Andrew C & Silomon, Jantje (2019). Dangerous Gaming: Cyber-Attacks, Air-Strikes and Twitter.


Working Paper

  • Dwyer, A. C., Hallett, J., Peersman, C., Edwards, M., Davidson, B. I. & Rashid, A (2022). How darknet market users learned to worry more and love PGP: Analysis of security advice on darknet marketplaces.
  • Patnaik, Nikhil, Dwyer, Andrew C, Hallett, Joseph & Rashid, Awais (2021). Don't forget your classics: Systematizing 45 years of Ancestry for Security API Usability Recommendations.