Call for Participation

Jamillah Knowles & Digit / https://betterimagesofai.org / https://creativecommons.org/licenses/by/4.0/

Core Questions

Within the scope of the HCOMP 2026 conference, shaping effective human oversight architectures represents both an urgent responsibility and a significant research challenge. Once regulatory mandates are fully in force, system designers and practitioners must move beyond basic individual user interface paradigms to address these core questions:

How can we design distributed systems that effectively communicate AI reasoning to support diverse human overseers?
How can we build scalable human oversight architectures that provide humans with both the structural understanding and the genuine causal influence and/or agency to control complex AI systems effectively?
How can we ensure that overseers are guarded from oversight fatigue?
How can human feedback, crowdsourcing, and preference learning be dynamically integrated to continuously detect, communicate, and repair human-AI misalignment at runtime?
How can human oversight minimise or prevent risk to health, safety, or fundamental rights that may emerge when a high-risk AI system is used? [Art. 14(2) AI Act]
How can human oversight be commensurate with different levels of risk? [Art. 14(3) AI Act]
For high-risk AI systems how can meaningful verification and confinement by at least two natural persons with the necessary competence, training and authority be designed? [Art. 14(5) AI Act]

Workshop Format

Lightning Introductions: At the start of the workshop, and to serve as an icebreaker, participants respond to a few questions, introducing themselves and their interests---(a) Who are you? (b) What is one fun thing about you that we probably do not know? (c) What would you like to get out of this workshop?
Paper Session: Authors of accepted papers can pitch their work through lightning talks, followed by Q&A and discussion. This setting encourages engagement between authors and other attendees.

Yutong Liu & Digit / https://betterimagesofai.org / https://creativecommons.org/licenses/by/4.0/

Breakout Sessions: These serve the goal to enable small group discussions on specific topics. We aim to create space for more active participation and diverse perspectives, which are then brought back into the main plenary.
Panel Discussion with Invited Speakers: Informed by the discussions from the breakout sessions, the panel will revisit critical questions at the intersection of their disciplines and invite further engagement from attendees.

Submission Details

As AI systems deploy in high-stakes domains, from healthcare to automated platform governance, algorithmic failures directly threaten human safety and rights. While emerging mandates like the European AI Act enforce a "human-in-the-loop,'' (Article 14 AI Act), effective oversight remains a profound socio-technical challenge. Poorly designed systems risk creating a dangerous illusion of control, leaving operators with nominal liability but zero practical agency.

Aligning with HCOMP 2026's focus on Human-AI Complementarity and Alignment and the conference theme of "Connections,'' this workshop unites researchers and practitioners across AI, human computation, HCI, psychology, law, and policy. Together, we aim to usher human oversight from a regulatory checkbox into a rigorous science of distributed monitoring and scalable human intervention.

We invite contributions addressing a broad range of relevant topics, including:

Distributed Oversight: Coordinating workflows across operators, domain experts, and compliance officers.
Scalable Steering: Interaction paradigms that alleviate cognitive fatigue, oversight burnout, and automation bias.
Misalignment Repair: Real-time mechanisms to detect, communicate, and fix human-AI misalignment at runtime.
Evaluation and Policy: Translating legal frameworks into verifiable software architectures and human-centric metrics.

We welcome full research papers (4 to 6 pages excluding references) and short position papers (2 to 4 pages excluding references) using the standard ACM format. Follow the submission templates of the HCOMP 2026 conference.

Submissions will be processed through a lightweight peer-review, and accepted works will be featured in interactive presentations and collaborative breakout sessions. Papers are non-archival to encourage interdisciplinary dialogue and works-in-progress.

Please submit your contributions as a single PDF by emailing the organizers at: S.Biswas@tudelft.nl, jiyounj@andrew.cmu.edu, m.t.sekwenz@tudelft.nl, mhlee@smu.edu.sg, harmank@umn.edu, simo.hosio@oulu.fi, u.k.gadiraju@tudelft.nl.

Important Dates

All deadlines below are per AoE (anywhere on earth time zone).

Paper submission Deadline : July 31, 2026
Notification to Authors: August 7, 2026
Conference Early-bird Registration: August 22, 2026
Workshop Date: Sunday, September 27, 2026

Page updated

Google Sites

Report abuse