Knowledge, Skills, Attitudes, Production: Competency-Based Education After Generative AI

Synthesis: This conceptual paper proposes adding production — the capability to deliver professional-standard work by directing tools and other people — as a fourth attribute of competency-based education (CBE), alongside knowledge, skills, and attitudes/values. The proposal responds to a construct-validity problem: generative AI has severed the inference from a student-produced artifact to the student's own knowledge and skill, and production supplies the missing interpretation for the "open lane" of two-lane assessment architectures.

Key Findings

1. A construct-validity crisis. GenAI severs the inference from a student-produced artifact to the student's own knowledge and skill, whether or not any policy was violated. Detection is unreliable, prohibition is unenforceable, and redesigning assignments to resist GenAI is an arms race — none of which answers what a GenAI-integrated artifact is evidence of.

2. Production as a fourth attribute. Production is defined as the capability to deliver work of professional standard by directing tools and other people, with three components: direction, evaluation and integration, and warranting.

3. Complements the two-lane architecture. Production supplies the missing interpretation for the "open lane" of two-lane assessment, where secure conditions certify knowledge and skills while the open lane assesses production.

Implications

For Assessment and Assessment Validity in the Generative AI era, the paper reframes the goal from policing artifacts to teaching and evaluating students' ability to direct AI toward professional-standard output. This connects to Psychometrically Aware AI and to Academic Integrity debates, moving beyond detection toward a coherent account of what competence means when AI does much of the generating.

The two-lane framing (secure conditions vs. open lane) is directly relevant to Automated Assessment and to institutional Curriculum Design decisions in Higher Ed and professional education. As a conceptual preprint, the framework would benefit from empirical operationalization of production and its assessment criteria.

Connected Concepts

Connected Articles

Citation

Hughes, J. W. (2026). Knowledge, skills, attitudes, production: Competency-based education after generative AI. EdArXiv preprint.