From Primary Education to Premium Workforce: Drawing on K-12 Approaches for Developing AI Literacy

University HCI research team with national trade union partner

Denmark

Researchers facilitating workshop; union organizing participation

• Private program

Full-day facilitated workshop (workplace-oriented AI literacy)

One intensive day plus three-month follow-up survey

53 participants (pre/post materials and follow-up)

Hands-on ML exploration via ml-machine.org; presentations and group discussions

• Single-country context with high baseline digitalization
• Short intervention may be insufficient to shift efficacy and empowerment
• Generalization to other unions or sectors requires further trials
• Self-report measures may reflect perceived rather than demonstrated competence

Adult professionals (union members / workforce cohort)

Mixed prior workplace exposure to ML applications

Not required; workshop uses accessible ML authoring and discussion scaffolds

• Increase understanding of core ML ideas relevant to workplace technologies
• Practice critically analyzing proposed ML systems using DORIT-style questioning
• Connect CCI-derived pedagogies to adult upskilling contexts

• Stimulate peer discussion about ML potentials and harms in participants' work settings
• Seed ambassadorship and union-mediated communities of practice

• Not a randomized multi-site trial
• Not a programming-intensive deep ML course
• Not primarily about generative LLM workplace policies

ml-machine.org ML authoring / exploration tool plus facilitated critique models

• Co-creator
• Evaluator

Danish professional context (materials likely Danish/English mix per venue)

• Short inputs introducing ML fundamentals
• Embodied and discussion-based group work using CEML and DORIT
• Participants produce workplace ML system concepts and analyses

• Critical examination of bias, accountability, and fairness in workplace ML scenarios
• Consent for non-anonymized appearance in publication (as stated in paper)
• Avoid overpromising empowerment from one-day events without infrastructural follow-up

• Greetings and framing of workplace AI/ML stakes
• CEML-grounded ML fundamentals and ml-machine.org hands-on block
• DORIT-guided critique of technological systems
• Group discussions on ML potentials and challenges; synthesis

• Participants retain judgment on workplace adoption; tools illustrate mechanisms
• Facilitators steer epistemic quality and connect examples to union concerns

• Import K-12-tested models (CEML, DORIT) to lower intimidation for adults
• Use tangible artifacts and structured vocabularies to externalize values and risks

• Knowledge increased while self-efficacy and empowerment scales did not
• Suggests long-horizon support, communities, and ambassadorship beyond one-shot training
• Adults outside formal education lack scaffolding ecosystems available to K-12 CCI projects

• Recontextualized CCI methods from classrooms to union-organized professional development
• Combined survey battery inspired by computing education research with qualitative artifact coding

• Participants produced substantive workshop artifacts (ML system ideas and analyses)
• Follow-up survey captured persistence of perceptions three months later

• Statistically significant improvement in self-reported ML knowledge pre to post
• No significant improvement in self-efficacy or empowerment—aligned with prior CCI empowerment measurement challenges

Authors argue HCI should invest in participatory infrastructuring with unions and AI literacy ambassadors to sustain empowerment beyond initial knowledge gains.

• Workplace ML examples may involve sensitive operational data; anonymize scenarios in facilitation
• Survey data require GDPR-aligned handling and clear union–research data agreements
• Non-anonymized consent for photos or quotes must be freely given and documented
• Discuss algorithmic discrimination using realistic but non-stigmatizing cases

• Post assessment
• Activity documentation
• Practitioner observation

• Longitudinal designs pairing workshops with ambassador networks and microcredentials
• Objective knowledge checks complementing self-report surveys in adult AI literacy

• Workplace AI literacy
• Participatory design with labor organizations
• CCI methods transfer to adult learning

• Completed

Adults

Union-organized professional learning

ML literacy + critical sociotechnical analysis

Workshop (CCI-adapted)

Medium

Medium (workplace examples, surveys)