AI literacy glossary
Key terms from this course, in plain English.
- 4D Framework
- Anthropic's AI fluency framework of four competencies: Delegation, Description, Discernment and Diligence.
- Agent (AI agent)
- An AI system, usually built on a large language model, that can plan and take actions with tools — such as searching, sending messages or running code — towards a goal.
- AI literacy
- The skills, knowledge and understanding that let people make informed use of AI and be aware of its opportunities, risks and possible harms. Organisations providing or using AI in the EU must take measures to support it among staff (EU AI Act, Article 4).
- Algorithm
- A set of steps a computer follows to solve a problem. Machine learning algorithms learn their behaviour from data.
- Anonymisation
- Removing or altering information so that people can no longer be identified. Truly anonymised data is outside data protection law; data where identifiers are only replaced (pseudonymised) is not.
- Artificial intelligence (AI)
- Computer systems that perform tasks associated with human intelligence, such as recognising images, understanding language or making predictions.
- Automation bias
- The tendency to over-trust suggestions from an automated system, even when there's evidence they're wrong.
- Bias
- Systematic errors or unfairness in AI outputs, often inherited from training data or design choices.
- Chatbot
- A program that holds a conversation with users in text or speech. Modern chatbots are usually built on large language models.
- Classification
- An AI task that sorts inputs into categories, such as spam or not spam.
- Computer vision
- AI that interprets images and video.
- Deep learning
- Machine learning using neural networks with many layers. It powers most recent AI breakthroughs.
- Deepfake
- Realistic AI-generated or manipulated image, audio or video that falsely appears authentic, often of a real person.
- Enterprise AI
- Business versions of AI tools provided under contracts that typically set out data handling, security and retention commitments.
- Generative AI
- AI that creates new content — text, images, audio, video or code — based on patterns learned from training data.
- Hallucination
- Plausible-sounding AI output that is false, invented or not supported by the source material.
- Human oversight
- People monitoring AI systems with the understanding, time and authority to question and override them.
- Large language model (LLM)
- A generative AI model trained on vast amounts of text to predict the next piece of text. The basis of ChatGPT, Claude, Gemini and Copilot.
- Machine learning (ML)
- Building AI systems that learn patterns from data rather than being programmed with explicit rules.
- Multimodal
- Able to process or produce more than one type of data, such as text, images and audio.
- Personal data
- Information relating to an identified or identifiable living person. Protected by UK GDPR and the EU GDPR.
- Prompt
- The instructions, question and context you give a generative AI tool.
- Prompt injection
- An attack where hidden instructions in content an AI reads try to make it do something its user didn't intend.
- Shadow AI
- Use of AI tools at work that haven't been approved by the organisation, often through personal accounts.