ARCHED: A Human-Centered Framework for Transparent, Responsible, and Collaborative AI-Assisted Instructional Design

Humans remain in control: ARCHED keeps educators as the primary decision-makers rather than having AI automatically generate instructional materials, ensuring that teachers—not algorithms—make final choices about what students learn.

Transparency through structured workflow: The framework uses a "cascaded workflow" aligned with Bloom's taxonomy (a well-established educational framework), making it clear how AI recommendations are generated and why, rather than treating AI suggestions as a "black box."

Multiple perspectives improve quality: The system uses specialized AI agents—one generating diverse pedagogical options and another evaluating alignment with learning objectives—to provide educators with multiple viewpoints before deciding, rather than a single AI recommendation.

Pedagogical rigor over pure automation: Unlike systems that prioritize speed and automation, ARCHED prioritizes sound educational principles, ensuring that AI assistance actually improves learning outcomes rather than just reducing teacher workload.

Empirical validation of benefits: The paper demonstrates through testing that ARCHED both enhances instructional design quality and preserves educator oversight, suggesting this human-centered approach doesn't require sacrificing educational effectiveness.