Long papers
Up to 9 pages, excluding references and appendices.
NLP/CORE Unranked
HAIPS@COLM 2026
2nd Workshop on Human-Centered Privacy and Security for Language Models
HAIPS@COLM 2026 is the second workshop focused on human-centered privacy and security for language models, aiming to bridge the gap between the language modeling community and researchers in HCI, usable security, privacy, and policy. It seeks to ground technical advancements in real-world human experiences, addressing risks arising from how people interact with language models as assistants, tutors, and autonomous agents.
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Short papers
Up to 4 pages, excluding references and appendices.
Research areas in scope
01
Topics of Interest
Human expectations, consent, and control over data used in LM systemsPrivacy- and security-preserving LM infrastructure for protecting people and institutions in real-world deployment contextsUser-facing protections that help people understand and prevent LM privacy and security risksHuman oversight, delegation, and accountability in agentic, personalized, and tool-using LM systemsEvaluation and benchmarking of LM and LM-system privacy, security, and safety through a human lensHuman impacts of adversarial LM-enabled privacy and security threatsUser vulnerability, manipulation, deception, and dark patterns involving language modelsEffects of language models used as simulated users, evaluators, auditors, or research participantsLM-powered tools that support privacy, security, safety, and human agencyDeveloper and organizational practices, risks, and responsibilities in LM design and deploymentHuman-centered policy and governance for LM privacy, security, and safety
Policies worth checking twice
- Submissions must be double-blind and reviewed via OpenReview
- At least two reviewers per paper
- Conflicts of interest managed through OpenReview’s COI detection and self-declared conflicts
- Accepted papers are non-archival, allowing authors to submit revised versions to other venues
- Submissions may use COLM, NeurIPS, or EMNLP format
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026