CUI 2026 conference logo
SAPH-AI

SAPH-AI

Security, Agency, Privacy of Human-AI

SAPH-AI is a workshop on privacy, security, agency, identity, and governance in human-AI ecosystems, with a focus on conversational AI, generative systems, and agentic technologies.

ACM CUI 2026 July 21–24, 2026 Bremen, Germany Full-Day Workshop

About

SAPH-AI brings together researchers and practitioners interested in how security, agency, and privacy jointly shape human-AI interaction.

Conversational AI, generative systems, and autonomous agents increasingly process sensitive personal data, construct behavioral profiles, integrate multimodal signals, and act on behalf of users across digital environments. In doing so, they influence autonomy, decision making, and identity formation.

SAPH-AI examines privacy, security, and agency as interconnected foundations of human-AI ecosystems. The workshop focuses on privacy risks, identity inference, multimodal leakage, delegated authority, transparency, accountability, and privacy-centered design for conversational and agentic systems.

By bringing together researchers from HCI, security, privacy, AI systems, machine learning, design, and social computing, SAPH-AI aims to advance a forward-looking research agenda for secure and privacy-centered human-AI ecosystems.

Important Dates

Conference Information

Conference ACM CUI 2026
Conference Dates July 21–24, 2026
Venue Bremen, Germany

Workshop Timeline

Submission Deadline TBD
Notification to Authors TBD
Camera-Ready Deadline TBD
Workshop Day TBA
Conference dates and venue are fixed. Workshop-specific deadlines and the exact workshop day will be announced once finalized.

Call for Participation

SAPH-AI invites contributions from researchers and practitioners working at the intersection of conversational AI, privacy, security, HCI, AI systems, and governance.

Submission Formats

  • Short position papers (2 to 4 pages, excluding references)
  • Privacy-centered system prototypes
  • Design provocations and early-stage research artifacts
  • Conceptual, empirical, technical, policy, and design perspectives

Review Approach

  • Light peer review by organizers and invited reviewers
  • Evaluation of clarity, originality, interdisciplinarity, and relevance
  • Emphasis on discussion value, not only maturity of results
  • Space for emerging work and research in progress

Topics of Interest

  • Frameworks linking privacy, identity, and agency in human-AI systems
  • Privacy risks, behavioral inference, automation bias, and identity persistence
  • Privacy-by-design architectures and safeguards for conversational AI
  • Accountability, consent, governance, and cross-jurisdictional data stewardship
  • Selective disclosure, contextual integrity, memory inspectability, and secure delegation
  • Critical perspectives on disproportionate impacts for historically marginalized or vulnerable populations
Accepted contributors will be invited to present short talks, join structured discussions, and, where appropriate, demonstrate early-stage systems and prototypes.

Agenda

SAPH-AI is designed as a full-day interactive workshop balancing peer-reviewed research, privacy-centered system prototypes, collaborative design activities, and research agenda building.

08:45–09:00
Welcome and Framing
Opening remarks, shared vocabulary, and workshop framing around privacy, agency, identity, and delegated AI.
09:00–10:00
Keynote: Privacy, Agency, and the Future of Delegated AI
Invited keynote from a scholar or practitioner bridging conversational AI, usable security, privacy engineering, governance, or sociotechnical analysis of identity and technology.
10:00–10:20
Break
Short break.
10:20–12:00
Paper Session I: Privacy Risks in Human-Agent Ecosystems
Short presentations and moderated discussion on agent-mediated data exposure, identity inference, behavioral profiling, persistent identity representations, cross-modal leakage, and privacy harms affecting vulnerable populations.
12:00–13:30
Lunch
Informal networking and exchange.
13:30–14:30
Paper Session II: Security, Governance, and Accountability
Presentations on defense-by-design architectures, secure coordination and delegation, transparency, dynamic consent, institutional governance, and cross-jurisdictional data flows.
14:30–14:45
Break
Short break.
14:45–15:30
Prototypes for Privacy-Centered Agentic Systems
Design explorations and early-stage systems embedding data minimization, identity protection, selective disclosure, memory inspection, auditability, and structural safeguards.
15:30–16:15
Collaborative Breakout Sessions
Interdisciplinary group work on privacy as a precondition of agency, identity persistence, privacy by design, and governance in multi-agent ecosystems.
16:15–16:30
Break
Short break.
16:30–17:00
Synthesis and Research Roadmapping
Shared taxonomy building, synthesis across themes, priority research directions, and discussion of follow-up collaboration, publications, and future workshop continuity.

Accessibility and Inclusion

SAPH-AI is committed to fostering an accessible, respectful, and inclusive workshop environment.

Accessible Materials and Participation

  • Accessible workshop materials, alt text, and readable formatting
  • Clear language and advance sharing of materials where feasible
  • High-contrast presentations and accessible slides
  • Support for multiple participation modes, including structured discussion and written contribution

Inclusive Community Practice

  • Attention to early-career researchers and scholars from underrepresented communities
  • Respectful dialogue and active moderation
  • Alignment with ACM and SIGCHI codes of conduct
  • Post-workshop continuity through summaries, shared documents, and follow-up collaboration
SAPH-AI recognizes that privacy research often intersects with lived experience, structural inequality, and sensitive topics, and the workshop will be designed to support equitable and meaningful participation.

Organizers

Sanchari Das George Mason University

Assistant Professor in the Information Sciences and Technology Department at George Mason University, where she leads research at the intersection of usable security, privacy, and human-centered AI. She directs the Data Agency & Security Lab and has extensive experience in interdisciplinary privacy and security research.

Rahat Masood UNSW Sydney

Senior Lecturer in the School of Computer Science and Engineering at UNSW Sydney and Director of the UNSW Online Cyber Security Program. Her work focuses on privacy, security, responsible AI, privacy-preserving systems, behavioral biometrics, misinformation, and risks from emerging AI.

Siddharth Das Sarkar George Mason University

Incoming PhD student in Information Science and Technology at George Mason University and current Machine Learning Specialist at Victaulic. His interests include applied AI, cybersecurity, authentication systems, identity verification, and secure AI-driven infrastructures.

Dakuo Wang Northeastern University

Associate Professor at Northeastern University, jointly appointed in Khoury College of Computer Sciences and the College of Arts, Media, and Design. His research lies at the intersection of human-computer interaction and artificial intelligence, with a focus on human-centered AI systems.

Filipo Sharevski DePaul University

Associate Professor of human-centered cybersecurity at DePaul University and director of the Actionable Cybersecurity & Accessibility Lab. His work focuses on usable security, misinformation, accessibility, and equitable cybersecurity education.

Salil Kanhere UNSW Sydney

Professor in the School of Computer Science and Engineering at UNSW Sydney, where he leads the Information Security and Privacy research group. His research spans networked systems, cybersecurity, and artificial intelligence, with extensive leadership across IEEE and ACM venues.