Lorin Schöni
ETH Zurich
HATS Workshop
August 23, 2026 | Room C109 | Starts 13:00
How do I get there?Workshop on Designing Human-AI Teaming for Security @ SOUPS 2026
Exploring how to balance human agency and AI autonomy in high-stakes cybersecurity decision-making.
Workshop Topic & Goals
The Workshop on Designing Human-AI Teaming for Security (HATS) addresses a central challenge for modern cybersecurity: as AI becomes more agentic, guaranteed human oversight is eroding.
HATS brings together researchers and practitioners to examine how human-AI teaming can strengthen security operations while preserving meaningful human agency in decision-making.
Objectives
- Co-create design principles for human-AI teaming in cybersecurity decision-making.
- Formulate actionable interaction design recommendations for research and practice.
- Build interdisciplinary networks that continue beyond the workshop.
Program
The workshop takes place on Sunday afternoon, August 23, 2026 in Room C109.
We follow an open door policy until capacity is filled. Priority is given to authors and participants who have signed up.
Coffee and drinks are available from 12:30. The workshop starts at 13:00. The break is from 15:00 to 15:30.
Detailed programme
- 13:00-13:15 Welcome and introduction
- 13:15-14:00 Presentations and questions
- 14:00-14:30 Ice breakers and group formation
- 14:30-16:15 Interactive group work (includes the 15:00-15:30 break)
- 16:15-17:00 Vernissage and closure
Accepted Papers
Download the workshop style file hats2026.sty. If you need a workshop-only ticket, please contact soups26chairs@googlegroups.com.
User-Agentified Security
Filipo Sharevski (DePaul University)
(Open paper (PDF))
Keeping Humans in Practice: Skill Atrophy as a Design Constraint for Human-AI Teaming in Cybersecurity
Tarini Saka (Max Planck Institute for Security and Privacy)
(Open paper (PDF))
The State of Human (Un) Centered AI Security Research
Leandra Willi, Rebecca Balebako (Zurich University of Applied Sciences (ZHAW))
(Open paper (PDF))
Application of Explainable AI in Identifying Fraudulent Transactions
Guanming Shi (Max Planck Institute for Security and Privacy)
(Open paper (PDF))
Contestable Reasoning Structures for Human-AI Teaming in SOC Investigations
Despoina Giarimpampa (University of Luxembourg); Roland Meier (Cyber-Defence Campus, armasuisse); Tegawendé F. BISSYANDE, Vincent Lenders, Jacques Klein (University of Luxembourg)
(Open paper (PDF))
Human-AI CTI Triage: Boundary Policies for Safe and Efficient Analyst Workflows
Jan Henri Beyer, Philipp Kühn, Markus Bayer (PEASEC, TU Darmstadt)
(Paper embargoed)
Error Taxonomy and Recovery Patterns for AI-Augmented CTI Pipelines
Philipp Kühn, Markus Bayer, Jan Henri Beyer (PEASEC, TU Darmstadt)
(Paper embargoed)
Position: Human Factors Reshape Adversarial Analysis in Human-AI Decision-Making Systems
Shutong Fan, Lan Zhang, Xiaoyong (Brian) Yuan (Clemson University)
(Open paper (PDF))
Call for Participation
We invite short position papers and contributions related to human-AI teaming in cybersecurity.
This is an open-door workshop until capacity is filled. Preference will be given to authors and participants who have signed up.
- Submission deadline:
June 1, 2026 - Notification to authors:
June 15, 2026 - Camera-ready deadline:
June 25, 2026
To participate, please submit a two- to four-page position paper on decision-making and human-AI teaming in cybersecurity. Submissions must follow the SOUPS formatting template. Submissions should not be anonymised. At least one author of each accepted paper must attend in person. Accepted papers will be published as workshop proceedings.
Workshop-only tickets: If you are interested, please contact soups26chairs@googlegroups.com.
Style file: Download the workshop style file hats2026.sty.
Topics of Interest
Example themes include (but are not limited to):
- Designing appropriate automation levels and agentic AI in cybersecurity
- Automation bias, trust calibration, and employee perceptions of AI decisions
- Human-centric frameworks and workflows for shared security decision-making
- AI-augmented threat hunting, signal processing, and SOC triage
- Technical, psychological, interpersonal, and societal risks in human-AI security workflows
- Error handling in AI-driven security tools and organizational implications
- Deployment barriers and facilitators for human-AI collaboration in enterprise security
- Human-AI teaming opportunities to address cybersecurity workforce shortages
Organizers
For any questions: contact hats-soups26@googlegroups.com
Alexandra von Preuschen
CISPA
Neele Roch
ETH Zurich
Tarini Saka
Max Planck Institute for Security and Privacy
Luisa Jansen
University of Bern
Marlene Wagner
ETH Zurich
Adrienn Toth
ETH Zurich
Steffen Holter
ETH Zurich
Noé Zufferey
ETH Zurich
Verena Zimmermann
ETH Zurich
Getting There
Locate the horse in front of the main building, enter the building through the main stairs for registration, then turn right - the first door you’ll see is our room. Don’t forget to grab some coffee and fruits before!
