More than prompting, more than ethics
When people ask “are you AI literate?”, they often mean “can you write effective prompts?” That’s like asking if someone is information literate and only checking whether they can use Google. AI literacy cannot be reduced to technical knowledge, operational skill, ethical awareness, or critical thinking alone—it requires integration across all of these, with particular emphasis on developing contextual judgement.
AI literacy
One-sentence definition: The multidimensional capability to recognise AI systems, critically evaluate their outputs and limitations, use them effectively for appropriate purposes, create meaningful outcomes through collaboration with AI, understand ethical implications of AI engagement, and develop contextual judgement about when and how AI serves professional goals.
The mistake in most AI literacy discussions is treating it as a single capability. Can you prompt? Can you spot bias? Do you understand how LLMs work? These questions matter, but answering any one of them tells you almost nothing about someone’s actual AI literacy. A person can be technically sophisticated about how transformers work yet have terrible judgement about when to use AI. Someone can write elegant prompts yet accept outputs uncritically. Another might understand ethics deeply but lack the operational competence to actually engage productively with AI tools.
This is why AI literacy applies a common architecture of literacy to artificial intelligence. Like information literacy or digital literacy, it’s not a single skill but a constellation of capabilities that must work together. The framework identifies six dimensions, each necessary but none sufficient.
The six dimensions
Access and recognition means understanding what AI systems are and recognising when they’re present and relevant. This includes distinguishing AI from other technologies, understanding AI as a language-based cognitive extension rather than a tool, search engine, or database, and identifying appropriate contexts for engagement. An AI-literate person knows when AI might help, when it’s likely to hinder, and when it’s simply irrelevant.
Critical evaluation involves assessing AI outputs for quality, accuracy, and reliability. AI systems can hallucinate. They produce fluent responses that sound authoritative but may be completely wrong. They exhibit bias in subtle and not-so-subtle ways. Understanding complementary errors—where humans and AI make different kinds of mistakes—matters as much as understanding AI limitations alone. An AI-literate person doesn’t accept AI outputs uncritically but maintains appropriate scepticism while remaining open to genuine assistance.
Functional application is the practical ability to use AI systems effectively for specific purposes. This includes operational competence with prompting, understanding how to structure requests for better outputs, knowing which AI tools suit which tasks, and deploying AI capabilities to achieve goals across domains. An AI-literate person can get useful work done with AI systems—but knows this is insufficient without the other dimensions.
Creation and communication moves beyond consumption to production—using AI to develop arguments, decompose complex problems, generate research questions, produce scholarly outputs, and communicate ideas. An AI-literate person creates through partnership with AI rather than merely receiving its outputs.
Ethical awareness and responsibility means understanding the social, ethical, and professional implications of AI engagement. This includes recognising issues of transparency and attribution, understanding when AI use might undermine rather than support goals, maintaining scholarly voice and integrity, avoiding over-reliance, and considering how AI engagement affects others. An AI-literate person makes responsible choices about when and how to use AI.
Contextual judgement and metacognition involves developing taste and professional judgement about meaningful AI engagement. This means distinguishing genuine competence from mere familiarity, recognising when AI helps versus hinders, evaluating both outputs and processes, and developing domain-specific judgement that evolves with experience. An AI-literate person knows not just how to use AI but when its use is appropriate and valuable.
Why this matters
AI literacy is neither binary (literate or illiterate) nor universal. It’s developmental and contextual. People begin where they are and progress through increasing sophistication. What constitutes meaningful AI engagement differs across domains—someone can be highly AI-literate in one context while still developing literacy in another.
The research confirms this multidimensional view. Long and Magerko’s 2020 framework identifies 17 competencies spanning recognition, understanding, evaluation, and critical engagement. UNESCO’s 2024 framework emphasises knowledge, skills, and attitudes working together. Recent frameworks increasingly stress that AI literacy involves moving from consumer to interpreter to collaborator—a progression requiring all six dimensions working in concert.
The practical implication: developing AI literacy cannot focus exclusively on any single dimension. Technical training without ethical awareness is dangerous. Ethical awareness without functional competence is impotent. Functional competence without critical evaluation is reckless. You need all six, integrated and developed together, appropriate to your domain and purposes.
Sources
- Allen, L. K., & Kendeou, P. (2023). ED-AI Lit: An Interdisciplinary Framework for AI Literacy in Education. Policy Insights from the Behavioral and Brain Sciences, 23727322231220339. https://doi.org/10.1177/23727322231220339
- Association of College & Research Libraries. (2016). Framework for information literacy for higher education. Association of College & Research Libraries. https://www.ala.org/sites/default/files/acrl/content/issues/infolit/framework1.pdf
- Long, D., & Magerko, B. (2020). What is AI Literacy? Competencies and Design Considerations. Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems, 1–16. https://doi.org/10.1145/3313831.3376727
Notes
The six-dimension framework isn’t the only way to conceptualise AI literacy, but it effectively captures what the research literature converges on: AI literacy is irreducibly complex, requiring integration across multiple capabilities that must develop together.