The ongoing wave of progress in AI has ushered in a transformative period in teaching and learning practices in higher education, offering opportunities and innovative tools while also bringing known and unknown drawbacks, risks, and threats.

AI literacy helps educators and learners understand how AI functions and what it means to use it responsibly, enabling them to make informed and safe choices. Based on frameworks by OECD, UNESCO, and the European Commission, LXS has designed an AI literacy framework addressing four key aspects of engaging with AI in teaching and learning.

Key Dimensions of Critical AI Literacy

• Understanding how AI systems function — statistically, not consciously
• Recognising the limitations and biases embedded in AI outputs
• Making ethical, responsible, and environmentally aware decisions about AI use
• Applying AI tools creatively and strategically in academic contexts

Artificial intelligence (AI) refers to machine-based systems that use input data to infer how to produce outputs — such as predictions, generated content, recommendations, or decisions — that can affect digital or physical environments.

Limitations of Generative AI

• Bias: GenAI can inherit and reproduce biases found in training data, including stereotypes and unbalanced perspectives.
• Hallucination: LLMs sometimes generate plausible-sounding but factually incorrect information — especially problematic in academic work.
• Lack of real-world understanding: AI systems lack common sense and genuine context comprehension.
• Outdated information: Responses are limited to the training data's knowledge cutoff — often out of date in fast-moving fields.

When used effectively and thoughtfully, AI can expand the boundaries of what is possible within your discipline. Following are approaches recommended by leading institutions including the Harvard Derek Bok Center for Teaching and Learning.

Seven Roles of AI in Teaching (Mollick & Mollick, 2023)

Simulations and Branching Scenarios

Use ChatGPT to design a 3-level decision tree for a branching scenario, then use a tool like Lovable to build an interactive responsive web application — no coding required. Students navigate realistic challenges and experience the consequences of their decisions in real time.

As AI becomes increasingly integrated into academic settings, demonstrating mastery now goes beyond completing traditional assignments. Faculty face a twofold task: updating current assessments to remain valid in an AI-rich environment, and identifying new ways to use AI constructively within assessment.

AI and Assessment Scale (Furze et al., 2024)

Examples of AI-Free Assessment

• Synchronous exams, oral exams, in-class writing, live presentations
• Context-based tasks tied to class discussions or local events
• Process-focused tasks requiring drafts and reflections showing thinking development

Examples of AI-Integrated Assessment

• Stasis Theory Exploration: AI guides students through structured argumentative questions; students reflect on AI's reasoning versus their own.
• Visualising Abstract Concepts: Students generate an AI image of a concept, write an interpretive essay, refine their prompt, and present both images.

Since we cannot rely solely on AI detectors to ensure integrity, it is vital to set clear expectations and transparent consequences for students. Having a clear and comprehensive syllabus or course AI policy makes this process efficient and effective.

Guidelines to Design Your Course AI Policy

• What are the non-negotiable learning outcomes that must be achieved without AI assistance?
• Where could GenAI act as a cognitive scaffold to help students reach higher-order thinking?
• How will you define original work in the context of this specific course?
• Which assignments are Red (No AI), Amber (AI as brainstormer/editor), or Green (Full AI integration)?
• If AI is permitted, what does proper attribution look like?

Sample Policy Statements

Ethical use of AI includes informed, thoughtful, and responsible engagement with AI systems. Critical AI literacy highlights that values, context, and responsibility are fundamentally intertwined with learning about and working alongside AI.

Key Ethical Considerations

• Amplification of bias: LLMs inherit biases from training data, potentially reproducing harmful stereotypes.
• Misinformation: Generative systems cannot reliably differentiate between accurate and false information.
• Copyright & Intellectual Property: Most LLMs are trained on data scraped without explicit creator consent.
• Environmental Impact: A standard conversation with an AI chatbot may consume up to 500ml of fresh water to cool servers (Li et al., 2023).
• Attribution & Licensing: Students and faculty should disclose not just that AI was used, but specifically how and with which tools.

Best Practices for Ethical AI Use

• Syllabus Integration: Establish clear expectations at the start of the course.
• Disclosure Statements: Require students to articulate how AI was used in each assignment.
• Developing AI Ethics as a Skill: Make critical reflection on AI's social, environmental, and economic impact a learning outcome.
• Informed Consent: Ensure students are aware of what data is collected by AI platforms.

Blended learning is defined as the "thoughtful integration of classroom face-to-face learning experiences with online learning experiences" (Garrison & Vaughan, 2008). It is an instructional model that leverages the unique strengths of both synchronous and asynchronous environments.

In a well-designed blended course, the online component becomes a natural extension of the classroom, allowing for self-paced foundational learning that frees up in-person time for high-level application, collaboration, and deeper inquiry.

Benefits of Blended Learning

• Accessibility: Removes physical and geographical barriers, allowing students to navigate life's unpredictability without pausing academic progress.
• Flexibility: Empowers students with agency to choose between in-person, synchronous online, or asynchronous attendance.
• Universal Design for Learning (UDL): Students choose the modality that best fits their unique learning style.
• Active Multimodal Engagement: Uses digital tools to create meaningful collaboration across modalities.
• Preparation for the Hybrid Workforce: Mirrors the modern professional world.

Synchronous vs. Asynchronous Activities

Example: Integrated Module on AI Ethics

• Asynchronous (Before): Students watch a 15-minute video on the history of AI ethics and read two contrasting case studies.
• Synchronous (During): Students meet in real-time for a Moot Court debate arguing the ethics of a specific AI application.
• Asynchronous (After): Students post a 300-word synthesis to a shared blog and provide peer feedback.

Blended course design should be grounded in learning science, pedagogy, and evidence-based research — not technology for its own sake.

Backward Design in Blended Courses

• Stage 1 – Identify desired results: What should students know, understand, and be able to do?
• Stage 2 – Determine acceptable evidence: How will you know if students achieved the outcomes?
• Stage 3 – Plan learning experiences: Which activities, content, and tools will help students achieve the outcomes?

The Community of Inquiry (CoI) Framework

Deep learning occurs at the intersection of three essential presences:

Mayer's Principles of Multimedia Learning

• Coherence Principle: Exclude all extraneous words, pictures, and background music from learning materials.
• Segmenting Principle: Break content into 5–10 minute micro-lectures rather than a single long video.
• Modality Principle: Pair graphics with narration rather than with on-screen text to use both cognitive channels.
• Personalization Principle: Use conversational language ("I," "you," "we") to increase social presence and motivation.

Blended Course Design Checklist

• ✓Has significant instructor presence online throughout the course
• ✓Uses less than half of in-class time for one-way delivery by lecture
• ✓Devotes significant in-class time to active learning and interaction
• ✓Engages students every week in activities both in class and online
• ✓Explicitly integrates online and in-class content and activities
• ✓Has frequent low-stakes assessments at regular intervals with timely feedback
• ✓Makes possible a degree of student choice over pathways to learning outcomes

Active learning is defined as "a pedagogical approach that engages students in the process of learning through activities and/or discussion in class, as opposed to passively listening to an expert" (Freeman et al., 2014). This shift moves focus from instructional input (what the teacher says) to cognitive output (what the student does).

Evidence-Based Benefits

• Higher-order thinking: Encourages analysis, critical evaluation, and synthesis above mere recall.
• Metacognition: Enables students to think about their own thinking and identify gaps in understanding.
• Soft skills: Builds interpersonal communication and collaborative problem-solving abilities.
• Long-term retention: Active retrieval leads to significantly better retention than passive re-reading or listening.
• Equity: Theobald et al. (2020) found active learning disproportionately benefits students from underrepresented backgrounds.

Constructivism

Learning is an active, contextualised process of building knowledge. Students construct understanding by integrating new information with prior experience. When new concepts conflict with existing knowledge, students must accommodate or assimilate — and this intellectual friction is where deep learning occurs.

Zone of Proximal Development (Vygotsky, 1978)

The most effective learning happens in the space between what a student can do independently and what they cannot yet do. Active learning strategies — peer discussion, guided inquiry — scaffold students through this zone toward higher cognitive complexity.

Dale's Cone of Experience

Abstract experiences (reading, listening) produce less retention than concrete, purposeful experiences (simulations, role-playing, doing the real thing). Active learning moves students from the peak of the cone toward the base — from passive observation to active participation.

Planning & Design

• ✓Give yourself extra time — active learning usually takes longer than a standard lecture
• ✓Focus only on activities that directly help students understand content aligned with learning outcomes
• ✓Explain the pedagogical rationale to students before beginning any active learning activity
• ✓Give clear instructions before distributing materials or asking students to form groups
• ✓Keep directions visible on a slide or the board throughout the activity

During the Activity

• ✓Move around the room — don't stay at the front
• ✓Know when to pause and address common difficulties with the whole class
• ✓Vary group composition — try randomised groupings rather than the same partners every time
• ✓Add a moment of silence after questions before students begin discussing

Managing Groups & Feedback

• ✓Let students know that any group member may be called upon to share
• ✓Assign specific roles (spokesperson, timekeeper, recorder) within groups
• ✓Save time at the end to review correct answers and key takeaways
• ✓Check for understanding with a follow-up individual question or poll

Research in learning science shows that courses designed with clear outcomes, aligned assessments, and purposeful activities produce stronger and deeper transferable learning than content-coverage approaches.

Hattie's (2009) synthesis of over 800 meta-analyses identified course design factors — clear learning intentions, structured feedback loops, and spaced practice — among the highest-impact influences on student achievement. Alignment between goals, instruction, and assessment is a prerequisite for expert-level student learning (Ambrose et al., 2010).

Backward Design (Wiggins & McTighe, 2005)

Also known as Understanding by Design (UbD), this framework inverts the traditional planning sequence:

• Stage 1 – Identify desired results: What should students know, understand, and be able to do?
• Stage 2 – Determine acceptable evidence: How will you know if students achieved the outcomes?
• Stage 3 – Plan learning experiences: What activities and tools build toward the outcomes?

Integrated Course Design (Fink, 2013)

L. Dee Fink's ICD shares Backward Design's emphasis on alignment but adds situational factors and extends learning goals beyond knowledge acquisition. Fink's six dimensions of significant learning:

Learning outcomes describe observable student behaviour, not instructor intentions. "Students will understand X" is difficult to assess. "Students will be able to analyse X and justify their interpretation using evidence" describes a demonstrable performance.

Bloom's Revised Taxonomy

The ABCD Formula

• Audience (A): Who is the learner?
• Behaviour (B): What observable action will they perform? (Use a Bloom's verb)
• Condition (C): Under what context or constraint?
• Degree (D): To what standard of performance?

A well-designed syllabus is both a contractual document and a learning-centered communication tool. Evidence suggests it shapes expectations, motivation, and initial engagement.

Key Components of an Effective Syllabus

• Course rationale: Why does this course matter? A compelling rationale motivates students from day one.
• Learning outcomes: Clear, measurable outcomes in student-friendly language, framed as what students will gain.
• Schedule and sequencing: A logical sequence that builds complexity progressively.
• Assessment overview: Transparent explanation of how and why students will be assessed.
• Course policies: Written in an inclusive, humanising tone acknowledging that students have complex lives.
• Support and resources: Proactive signposting to academic support, wellbeing services, and accessibility accommodations.

• 1Audit situational factors: Who are your students? What do they already know? What are the institutional constraints?
• 2Draft 4–8 course-level learning outcomes using Bloom's Taxonomy verbs at a range of cognitive levels.
• 3Design assessments first: decide how students will demonstrate achievement before planning lectures or readings.
• 4Map learning activities that give students practice with the thinking required by assessments.
• 5Sequence the course so complexity builds progressively and earlier material is revisited in new contexts.
• 6Write the syllabus: translate your design into a student-facing document using warm, inclusive language.

Higher education classrooms are among the most diverse learning environments in the world. Yet traditional instructional designs are mainly built around an implicit "model student" who may not represent the full diversity of today's learners.

A lack of belonging is among the strongest predictors of student dropout — particularly for first-generation, racially minoritised, and low-income students. Conversely, students in inclusive learning environments demonstrate higher engagement, greater academic resilience, and stronger long-term outcomes (Hockings, 2010).

Key Principles of Culturally Responsive Teaching

• Validating students' cultural identities as assets rather than deficits
• Using examples, texts, and case studies that reflect diverse cultural contexts
• Building caring, respectful, and high-expectation relationships with all students
• Critically examining whose knowledge is centred in the curriculum

First Impressions & Classroom Climate

• Learn and correctly pronounce students' names. Invite students to share their preferred names.
• Frame the course as a community of learners, not an evaluative contest.
• Use a warm, invitational tone in the syllabus and early communications.

Inclusive Classroom Norms

• Co-construct discussion guidelines with students so that norms of respectful engagement are owned collectively.
• Design participation structures that do not default to cold-calling or privilege the most verbally confident students.
• Acknowledge uncertainty and model intellectual humility to create psychological safety.

Instructor–Student Relationships

• Express genuine interest in students' goals, challenges, and perspectives.
• Maintain high expectations for all students equally. Low expectations are experienced as disrespect (Steele, 2010).

Diversifying Course Content

• Audit your reading list: whose perspectives are represented, and whose are absent?
• Use examples and case studies that reflect a range of cultural contexts.
• Where foundational texts were produced in historically exclusionary contexts, acknowledge this critically.
• Invite guest speakers whose backgrounds expand the range of voices students encounter.

Inclusive Assessment

• Offer choice in assessment format where possible — written, oral, visual, or practical.
• Provide detailed rubrics and exemplars in advance so all students understand expectations.
• Design low-stakes early assessment opportunities that allow students to receive feedback before higher-stakes tasks.

Inclusive Feedback Practices

• Provide feedback that is specific, actionable, and forward-focused.
• Avoid language that attributes performance to fixed ability rather than changeable behaviours.
• Offer feedback in multiple formats: written, audio, or in-person conference.

Course Design & Materials

• ✓Do my learning outcomes use accessible, student-friendly language?
• ✓Does my reading list represent diverse authorship, geography, and theoretical tradition?
• ✓Are all course materials available in accessible digital formats with captions and alt-text?
• ✓Does my syllabus use warm, invitational language that frames policies in terms of student success?

Assessment & Feedback

• ✓Do my assessments measure the intended learning outcomes?
• ✓Have I provided detailed rubrics and exemplars in advance?
• ✓Is my feedback specific, actionable, and framed in growth terms?

Classroom Environment

• ✓Do I know and correctly use students' preferred names?
• ✓Have I established inclusive discussion norms collaboratively with students?
• ✓Do I proactively signpost academic support and wellbeing resources?