AI for Computational Design Science: A Responsible Human-AI Framework and Case Study on Short-Form Video Safety Surveillance
A paper titled 'AI for Computational Design Science: A Responsible Human-AI Framework and Case Study on Short-Form Video Safety Surveillance' was published on arXiv (cs.AI) on 2026-09-04. It proposes AI4CDS, a five-phase methodological framework for computational design science where AI assists in problem formulation, resource construction, design search, evaluation, and knowledge abstraction, while researchers retain responsibility for domain grounding, admissibility, verification, and scientific judgment. The framework is instantiated through ChildRiskGuard, an interpretable artifact for detecting short-form videos inappropriate for children.
The paper introduces AI4CDS, a five-phase framework for integrating AI into computational design science research. It emphasizes responsible human-AI collaboration governed by graduated trust, reversibility, auditability, and differentiated reproducibility. The framework is demonstrated via ChildRiskGuard, an interpretable system for detecting short-form videos inappropriate for children, which separates generic from child-specific risk and represents distinct developmental-risk mechanisms.
The AI4CDS framework structures AI involvement across five phases: problem formulation, resource construction, design search, evaluation, and knowledge abstraction. It introduces governance mechanisms such as graduated trust, reversibility, auditability, and differentiated reproducibility. The ChildRiskGuard case study translates audience-dependent safety and explanation faithfulness into three technical challenges, developing an artifact that separates generic from child-specific risk and represents distinct developmental-risk mechanisms.
This research signals a growing emphasis on responsible AI integration in design science, particularly for safety-critical applications like children's content moderation. The framework's focus on auditability and human oversight may influence industry practices for AI-assisted research and development, especially in domains requiring high accountability.
The framework and artifact could support development of AI tools for content moderation, child safety, and responsible AI research workflows. Organizations may leverage AI4CDS to structure AI-assisted design processes with built-in accountability, reducing risk and improving trust in AI-generated designs.
Observable next signals include adoption of AI4CDS in other design science domains, further development of ChildRiskGuard or similar child-safety AI systems, and increased discussion of governance mechanisms for human-AI collaboration in research. Potential follow-up work may address scalability, generalizability, and empirical validation of the framework.