for Learning Professionals
MyFest25
August 19, 2025
Stella Lee, PhD.
AI Ethics Rubric
Two years ago, as part of a project with European Innovation in Technology (EIT), I developed an AI Literacy Framework. Of all the areas we explored, ethics was the one people talk a lot about but also often is treated as an after thought. �
Source: https://paradoxlearning.com/resource/ai-literacy-framework/
Question:
How would you describe your organization's current capability for making ethical decisions around AI use? �
Why AI Ethics in Education?
Ethics is no longer an afterthought. It is a design-level work.
“
”
The AI Ethics Rubric
Source: https://paradoxlearning.com/resource/ai-ethics-rubric-for-ld-a-design-tool/
Ten dimensions of focus:
Purpose & Value Alignment
Looks at whether the use of AI is intentional, pedagogically sound, and aligned with the organizational core values and strategic goals. It emphasizes using AI to enhance—not replace—human-centered learning experiences and ensures that deployments are driven by meaningful learning outcomes rather than novelty or convenience.
Question to ask: What assumptions about productivity, efficiency, or control are baked into this tool?
Transparency & Explainability
Assesses whether learners and stakeholders are communicated with and understand what the AI does, how it works, and why it makes specific decisions or recommendations.
Question to ask: What happens when an AI tool makes a decision learners disagree with – can they trace the logic?
Bias & Fairness
Focuses on identifying and addressing bias in AI systems—whether embedded in training data, algorithms, or design choices—to prevent the marginalization or exclusion of any learner group. Fairness ensures that AI tools provide equitable access, representation, and outcomes for all learners, regardless of background, ability, or identity.
Question to ask: Do we checked whether AI outputs reinforce stereotypes or exclude certain people?
Data Ethics & Privacy
Ensures responsible handling of learner data—collecting only what’s needed, securing it properly, and being transparent about how it’s used, stored, and archived.
Question to ask: Is learner data anonymized, protected, and governed by a clear deletion policy?
Accountability & Oversight
Defines who owns ethical oversight for AI tools. Involves governance mechanisms, review cycles, and documentation of risks and remediation plans.
Question to ask: Who is responsible for reviewing, approving, and updating our AI tools?
Learner Agency & Consent
Informs learners when AI is in use, obtains explicit consent where appropriate, and provides meaningful ways for them to influence how AI shapes their learning experience. This includes the ability to opt out, adjust preferences, and question or override AI-driven decisions that impact their development or progression.
Question to ask: Are there any implicit pressures to engage with the AI tool?
Ethical Use of GenAI
Encompasses issues of copyright, authorship, accuracy, creativity, and responsible use of large language models (LLMs).
Question to ask: GenAI as a co-creator or as a shortcut?
Vendor Due Diligence
Focuses on ethical procurement practice —including how vendors are evaluated through contracts, documentation, and alignment with your values. This dimension ensures that third-party tools and providers uphold responsible practices across design, data use, and sustainability.
Question to ask: What is their roadmap (if any) for responsible AI use – not just current capabilities?
Inclusive AI Design
Focuses on how AI tools must consider diverse learner needs—across languages, abilities, contexts, and cultures. Testing for accessibility and relevance is essential.
Question to ask: How are we involving underrepresented voices in the testing or design process?
Environmental Sustainability
Acknowledges the environmental impact of using AI tools in learning, including the energy consumption, carbon footprint, and resource intensity of training, hosting, and running AI systems.
Question to ask: What energy demands does this tool introduce (e.g. compute-heavy features)?
Three Levels of Maturity
Source: https://paradoxlearning.com/resource/ai-ethics-rubric-for-ld-a-design-tool/
Dimension | Emerging | Developing | Leading |
Purpose & Value Alignment | AI is adopted without clear alignment to learning goals, organization’s values, or strategy. | AI tools are somewhat aligned with organizational as well as learning goals and values, but not consistently evaluated for fit or impact. | AI initiatives are purpose-driven, learner-centered, and fully aligned with organizational mission and goals. |
Transparency & Explainability | AI decisions and processes are opaque to learners and stakeholders. | Basic explanations are provided, but transparency is inconsistent or overly technical. | AI operations are transparent, with clear, accessible explanations for outputs and recommendations. |
Bias & Fairness | No formal process exists to identify or address bias in AI systems or data. | Initial efforts to reduce bias exist, such as limited testing or content review. | Bias is continuously monitored, with structured audits and inclusive design baked into AI development. |
Companion Guide
With examples/scenarios and guiding questions.
Source: https://paradoxlearning.com/resource/ai-ethics-in-practice/
Wrap-Up
Final Reflection:
What’s one ethical lens or practice you’ll bring back to your workplace?�
Thank you!