Research map / Large language models / AI and society
Using Case Studies to Teach Responsible AI to Industry Practitioners
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Describes an experience report of co-developing and delivering two-session interactive case-study workshops on Responsible AI to Meta practitioners, with pre/post self-assessments showing improved RAI understanding.
Problem
Industry practitioners need effective educational resources and methods to learn Responsible AI (RAI), and such training is scarce.
Approach
A stakeholder-first educational approach using interactive case studies with positive/negative handout pairs, stakeholder matrices, and practitioner-authored case studies, co-designed with an industry partner.
Key results
- Self-rated RAI understanding improved from a weighted average of 4.88 (n=38) to 6.94 post-workshop (reported as 4.88 pre vs 6.94 post).
- 100% of respondents said they were motivated to apply workshop content to their work and that the case study format was effective for understanding RAI.
- 77% of respondents reported very to substantially improved RAI knowledge; 77% said they were engaged by participant-to-participant interaction.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| deliver an interactive RAI workshop using case studies | Industry practitioners at Meta · RAI workshop | self-assessed understanding of RAI concepts and techniques: 4.88 -> 6.94; 4.84 -> 6.94 | Helped |
| use case studies and small-group hands-on activities | Interactive RAI workshop · RAI workshop, post-survey | reported engagement from participant interaction and hands-on activities: 77% | Helped |
| use the case study format to teach RAI concepts | Interactive RAI workshop · RAI workshop, post-survey | agreement that the format was effective for understanding RAI concepts: 100% | Helped |
| deliver the workshop content | Industry practitioners at Meta · RAI workshop, post-survey | reported improvement in RAI knowledge: 100% | Helped |
| deliver the workshop content | Industry practitioners at Meta · RAI workshop, post-survey | motivation to apply workshop content to own work: 100% | Helped |
Surprising findings
- Attrition was substantial: just over 50% of first-session participants returned for the second session.
Limitations
- Experience report limited to workshops at a single industry partner (Meta).
- Self-reported pre/post assessments with small post-survey sample (n=16) and no longitudinal measurement.
- Case studies could not use proprietary internal information about company systems.
Experimental setup
Datasets: housing ads delivery case study (constructed) toxic comment moderation case study (constructed)
Metrics: pre/post self-assessed RAI knowledge (weighted average ratin survey-reported engagement and motivation participant role distribution
Builds on: stakeholder-first training
Closest papers by meaning
- We Are AI: Taking Control of Technology (AAAI 2025 · similarity 0.89)
- Advancing Research on Equitable AI Education Through a Focus on Implementation: Insights from a Middle School Computer Vision Module Beta-Test (AAAI 2025 · similarity 0.85)
- Developing Chatbots for Sustainability: Experiential Learning in an Undergraduate Business Course (AAAI 2025 · similarity 0.81)
- Supporting AI Literacy Teaching Through the Development of Assessments for Classroom Use (AAAI 2025 · similarity 0.81)
- Position: Standardization of Behavioral Use Clauses is Necessary for the Adoption of Responsible Licensing of AI (ICML 2024 · similarity 0.80)
- Making Transparency Advocates: An Educational Approach Towards Better Algorithmic Transparency in Practice (AAAI 2025 · similarity 0.80)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.