What is AI literacy?
The Harvard Chan AI Glossary
In our classes, we often pause to explain the ideas behind AI and the language used to describe it. These posts grew out of those conversations. They’re for anyone who would like to understand AI better, with no technical background assumed.
AI literacy is the knowledge and judgment needed to understand, use and critically assess artificial intelligence. It includes recognizing what a system can do, asking sensible questions about its outputs and deciding when its use is appropriate. It is not the same as being able to build an AI system or write computer code.
A person can become skilled at using a chatbot without understanding why it sometimes invents an answer. Equally, someone can memorize the names of different AI methods without knowing how to check a generated response. Literacy brings practical use and understanding together, rather than treating either one as sufficient.
Consider asking an AI assistant to summarize a report. A useful first question is whether it actually received the report. Next, compare important statements with the original text. Check whether the summary leaves out a qualification or presents an uncertain finding as established. Those habits matter more than knowing a clever phrase to add to the instruction.
Some basic distinctions help. A model is the component that has learned patterns from examples. An application is the product around it, which may add searches, saved information and tools. A fluent answer from a model is not necessarily a checked answer from a source. Knowing which parts are involved helps you understand what the result establishes.
AI literacy also includes awareness of information handling. What are you sharing? Where is it going? Who is allowed to access it? These questions apply even when the task itself seems harmless. An easy-to-use interface does not remove the need to understand the consequences of supplying confidential material.
There is a social dimension too. AI systems can affect what information people encounter and how opportunities are allocated. Being able to question those effects is part of literacy, not something reserved for programmers. People should be able to distinguish evidence of benefit from a persuasive demonstration.
For a healthcare leader, literacy might mean looking at a proposed discharge or referral summary and asking: did the assistant receive the actual record, which source supports this statement, and is this a suggestion or an action already taken? It also means recognizing that a useful draft does not establish a dependable clinical service. Those questions can be asked without writing code. The aim is enough understanding to use the system deliberately, challenge its claims and recognize when clinical or technical expertise is needed.
Sources: Long and Magerko — What Is AI Literacy?; UNESCO — AI Competency Framework for Students.
If you would like to find out more about AI and improve your leadership skills, consider the AI in Health Care Certificate of Specialization from the Advanced Learning Academy at Harvard T.H. Chan School of Public Health. You can take an individual course or work toward the full certificate.
AI is a fast-moving field. This explanation reflects our class discussions and related material at the time of writing. Terminology, capabilities and tools may change.
For suggested additions or corrections, please email me at tpanch@hsph.harvard.edu.
Dr. Trishan Panch
Program Director, AI in Health Care
Harvard T.H. Chan School of Public Health — Advanced Learning Academy