TRUST

Autonomous systems, including AI agents, are rapidly reshaping software engineering across domains such as software development, robotics, and cyber-physical systems. These systems are increasingly capable of acting with high levels of autonomy, interacting with users, and operating in complex and often critical environments. As their autonomy grows, agents are no longer merely tools that assist human stakeholders, but are increasingly taking on roles akin to team members and, in more extreme cases, even managerial functions such as coordination, delegation, and decision-making. While recent research has largely focused on leveraging such systems to support software engineering tasks (i.e., coding agents for software engineering), considerably less attention has been devoted to how these systems themselves can be systematically engineered to be trustworthy and responsible. In particular, ensuring properties such as fairness, accountability, transparency, and robustness remains a key open challenge.

The 1st International Workshop on Trustworthy and Responsible aUtonomous SysTems (TRUST) aims to address this gap by focusing on software engineering principles, methodologies, and automated techniques for building and assuring autonomous systems. Rather than focusing solely on the application of AI agents to SE, TRUST aims to emphasize the need to engineer autonomous systems as first-class software systems, with explicit consideration of ethical, societal, and quality-related concerns. This need is becoming even more pressing as increasingly autonomous agents move beyond the role of assistants and begin to operate as active participants in socio-technical settings, taking on responsibilities akin to team members and, in some cases, managerial functions such as coordination, delegation, and decision-making. The workshop will explore approaches for designing responsible systems, including requirements, specifications, governance mechanisms, and human oversight structures, as well as automated solutions for assessing and enforcing trustworthiness through testing, monitoring, and continuous integration pipelines.

TRUST brings together researchers, practitioners, and tool developers from software engineering, artificial intelligence, and related disciplines to discuss emerging challenges, share novel ideas, and define future research directions. Participants will have the opportunity to present new methodologies, tools, and empirical studies, as well as reflect on practical experiences and open challenges in domains such as AI-based software assistants, autonomous driving, and intelligent cyber-physical systems. By fostering interdisciplinary discussion, the workshop aims to advance the foundations of software engineering for autonomous systems and to build a community around trustworthy and responsible autonomous software.

Accepted Papers

Call for Papers

Autonomous systems, including AI agents, are rapidly reshaping software engineering by enabling systems that can act, reason, and interact autonomously with minimal human intervention. As these systems become more capable, they are increasingly moving beyond the role of tools and assistants, taking on responsibilities more akin to team members and, in some cases, managerial functions such as coordination, delegation, and decision-making. While recent research has largely focused on leveraging such systems to support software engineering tasks (i.e., agents for software engineering), considerably less attention has been devoted to how these systems themselves can be systematically engineered to be trustworthy and responsible. Ensuring properties such as fairness, accountability, transparency, robustness, and safety in autonomous systems remains a fundamental and largely unsolved challenge, especially as these systems begin to assume roles that influence coordination, oversight, and decision-making within socio-technical environments. Existing approaches often focus on model-level improvements or one-shot evaluations, without providing systematic software engineering methodologies or automated processes to continuously assure these properties in real-world systems.

Motivated by these challenges, the TRUST workshop (International Workshop on Trustworthy and Responsible aUtonomous SysTems) aims to bring together researchers and practitioners to explore software engineering foundations, methodologies, and automated techniques for building and assuring trustworthy and responsible autonomous systems. In particular, the workshop emphasizes a shift from Agents for Software Engineering to Software Engineering for Autonomous Systems, focusing on how to design, verify, and maintain such systems as first-class software engineering solutions.

Topics of Interest

Topics of interest include, but are not limited to, the following:

• Engineering trustworthy and responsible autonomous systems: requirements engineering, specification, governance, and design principles for fairness, ethics, and accountability, and safe delegation of autonomy
• Fairness, bias, and ethical behavior in autonomous systems: detection, mitigation, and prevention of biases in agentic and autonomous settings
• Automated testing and verification: techniques for fairness testing, ethical validation, robustness analysis, and trustworthiness assessment
• Continuous assurance pipelines: CI/CD approaches for monitoring, validating, and enforcing responsible behavior in autonomous systems
• Runtime monitoring and adaptation: detecting and correcting undesirable behaviors during system execution
• Specification and contracts for autonomous systems: formal and semi-formal approaches to define expected trustworthy behavior and constraints
• Human-in-the-loop and oversight mechanisms: integrating human judgment, intervention, and escalation paths for autonomous systems acting as assistants, collaborators, or decision-making entities
• Explainability, transparency, and accountability: techniques to make autonomous systems understandable, auditable, and governable in organizational and socio-technical settings
• Empirical studies and benchmarks: evaluation frameworks, datasets, and metrics for trustworthiness and responsibility
• Tooling and infrastructure: frameworks, libraries, and platforms for engineering and evaluating autonomous systems
• Applications and case studies: autonomous systems in software engineering, DevOps, robotics, autonomous driving, and cyber-physical systems
• Socio-technical and regulatory aspects: compliance, standards, and impact of regulations (e.g., AI Act) on autonomous systems
• Experience reports: lessons learned from deploying or evaluating autonomous systems in practice
• Autonomous agents in team and managerial roles: methodologies, safeguards, and empirical studies for systems that coordinate work, allocate tasks, supervise processes, or otherwise act as team members or managers

The aim of this workshop is to provide a forum for researchers and practitioners to present and discuss novel methods and techniques for engineering, evaluating, governing, and continuously assuring trustworthy and responsible autonomous systems, including those that increasingly act as collaborators or coordinators in human-centered workflows.

We expect that the workshop will help to:

• Provide researchers with a comprehensive view of current challenges and solutions in engineering trustworthy autonomous systems, including systems that operate as collaborators or coordinators within teams
• Investigate how automated software engineering techniques can support fairness, ethics, and trustworthiness of autonomous systems
• Reinforce the foundational knowledge around software engineering for autonomous systems
• Identify new opportunities for addressing ethical, organizational, and societal challenges in modern automated software engineering landscapes, particularly where autonomous systems influence authority, coordination, and decision-making
• Propose new empirical methodologies, protocols, and metrics for assessing trustworthiness and responsibility
• Analyze real-world applications and challenges in deploying autonomous systems
• Develop a roadmap for future research directions in trustworthy and responsible autonomous systems
• Clarify the implications of autonomous systems acting not only as tools, but also as team members or, in some settings, managerial actors

Submission Process

• Submission through HotCRP: https://trust26.hotcrp.com/
• Template: \documentclass[sigconf,review,anonymous]{acmart}
• Full papers: 5–8 pages (including references), reporting original research contribution, industry insights, replication or negative results, or experience reports
• Short papers: 4 pages (including references), presenting visions, position papers, novel ideas, preliminary results, or experience reports
• Artifacts: Authors are encouraged to provide an anonymous link to supporting artifacts or justify their absence
• Compliance: Submissions must comply with ASE policies (double-blind review, plagiarism, human subjects, open science)

Review Process

In adherence to the main conference guidelines, all submissions will undergo a rigorous peer-review process. Each paper will be reviewed by at least three program committee members following a double-anonymous process.

Submissions will be evaluated based on the following criteria:

• Soundness: The extent to which the paper’s contribution addresses its research questions and is supported by a rigorous application of appropriate research methods.
• Significance: The extent to which the paper’s contributions can impact the field of software engineering, in research or practice, and under which assumptions.
• Novelty: The extent to which the contributions are sufficiently original with respect to the state-of-the-art.
• Verifiability and Transparency: The extent to which the paper provides sufficient information to understand how the proposed approach works, how data was obtained and analyzed, and how results can be independently verified or replicated.
• Presentation: The quality of writing, including clarity of exposition, proper use of language, absence of ambiguity, and adherence to formatting guidelines.

Reviewers will consider all the above criteria when evaluating submissions. The paper should clearly explain and justify its contributions.