3rd Workshop on
Software Quality Assurance for AI

Artificial Intelligence (AI) is becoming increasingly important in our world, and is being included in a large number of applications and technologies that we use daily. Many AI-enabled applications are produced by developers without proper training in software quality practices or processes. An AI-enabled system is a software-based system that comprises non-trivial AI components in addition to traditional software components.

As with any software system, AI-enabled systems require attention to software quality assurance (SQA) in general and code and design quality in particular. Agile development models enable companies to choose technologies to adopt in their systems at any development stage. Therefore, it is challenging to anticipate if a system, or a data pipeline used to develop AI, will produce high-quality models and high-quality results.

You are invited to submit:

• A research or experience full paper with 8 pages max. Papers describing the challenges, starting results, vision papers, or the experience papers from or in cooperation with the practitioners are encouraged.
• A short research or experience paper with 4 pages max. Paper describing initial research results for which there is not yet a complete evaluation.
• Position paper with 2 pages. Paper describing how software quality and AI research and practice will look at least ten years from now.

The paper submissions will follow the full double-blind review process as recommended by the SANER organizer committee. Each submitted paper will be reviewed by at least three PC members and will be assessed for their pertinence to the workshop theme, for the quality of the writing and for the level of interest that the expressed position could raise in the community.

Evaluation Criteria

• Relevance: Submission must be related and in line to the Call for Papers.
• Novelty: Is there sufficient originality in the contribution, and is it clearly and correctly explained with respect to the state of the art?
• Soundness: Are all claimed contributions supported by the rigorous application of appropriate research methods? Claims should be scoped to what can be supported, and limitations should be discussed.
• Significance: Are contributions evaluated for their importance and impact with respect to the existing body of knowledge? The authors are expected to explicitly argue for the relevance and usefulness of the research and discuss the novelty of the claimed contributions through a comparison with pertinent related work.
• Replicability: Is there sufficient information in the paper for the results to be independently replicated? The evaluation of submissions will take into account the extent to which sufficient information is available to support the full or partial independent replication of the claimed findings.
• Presentation: Is the paper well-structured, clear, and free of grammatical errors? Does it follow the submission guidelines?

Submission Format

All submissions must be in PDF format and conform to the IEEE Conference Proceedings Formatting Guidelines (title in 24pt font and full text in 10pt font). LaTeX users must use \documentclass[10pt,conference]{IEEEtran} without including the compsoc or compsocconf option. Additionally, papers must comply with the IEEE Policy on Authorship.

Authors must comply with the SIGSOFT Open Science Policy, which requires archiving data and artifacts in a permanent repository (e.g., Zenodo, not GitHub) to the extent ethically and practically possible, and including links in a Data Availability section in their manuscripts.

SQA4AI 2026 employs a double-anonymous review process. Do not include author names or affiliations in submissions. All references to the author’s prior work should be in the third person. Any online supplements, replication packages, etc., referred to in the work should also be anonymized. By submitting to SQA4AI 2026, authors agree to the ACM Policy and Procedures on Plagiarism, Misrepresentation, and Falsification. Papers submitted must not be published or under review elsewhere. The Program Chairs may use plagiarism detection software under contract to the ACM. If the research involves human participants/subjects, the authors must adhere to the ACM Publications Policy on Research Involving Human Participants and Subjects.