AI Glossary for HR Professionals: 50 Terms Explained Simply
AI conversations in HR are full of jargon. This glossary defines the fifty terms HR professionals meet most often, in plain language, with a note on why each matters for people work.
Short answer
The essential AI terms for HR professionals include artificial intelligence, machine learning, generative AI, agentic AI, large language model, prompt, hallucination, retrieval-augmented generation, natural language processing, training data, algorithmic bias, adverse impact, explainability, human in the loop, and regulatory terms such as high-risk AI system, provider and deployer. Understanding them helps HR evaluate vendors, govern AI use and explain decisions to employees.
Key takeaways
- Terms are grouped alphabetically; each definition is written for HR practitioners, not engineers.
- Regulatory terms such as provider, deployer and high-risk AI system determine who is responsible for what under the EU AI Act.
- Fairness terms such as adverse impact, bias audit and the four-fifths rule are central to AI in hiring.
- Linked guides go deeper on the most important concepts.
Guides in this topic
- Common AI Terms Explained Simply: A Guide for HR Teams
Fifteen core AI terms with everyday analogies, HR examples and commonly confused pairs.
- What Is Machine Learning in HR? How It Works, Uses and Limits
How machine learning learns from people data, where HR uses it, and its limitations.
- What Is an LLM? Large Language Models Explained for HR
How large language models work, why they hallucinate, and how HR can use them safely.
Start with the three most important concepts, each with its own guide: common AI terms explained for a guided tour of the core vocabulary, what is machine learning in HR, and what is an LLM.
A
Adverse impact
A substantially different selection rate for a protected group compared with others in a hiring, promotion or other employment decision. It is a key measure when testing AI selection tools for fairness. Learn more: AI hiring bias.
Agentic AI
AI that pursues a goal by planning and carrying out multi-step tasks through connected software tools, with limited human supervision. In HR, agents can schedule interviews or run onboarding logistics. Learn more: Agentic AI in HR.
AI agent
A software system built on an AI model that can take actions, such as sending messages or updating records, to complete a task. An AI agent is the practical form of agentic AI. Learn more: What is agentic AI in HR.
AI literacy
The knowledge and skills people need to use AI effectively and responsibly, including understanding its capabilities, limits and risks. Under the EU AI Act, as amended in 2026, providers and deployers must take measures to support the development of AI literacy among their staff. Learn more: AI literacy in the workplace.
Algorithm
A defined set of steps or rules a computer follows to solve a problem or produce an output. Machine learning algorithms learn their rules from data rather than having them written by hand.
Algorithmic bias
Systematic errors in an AI system's outputs that unfairly favour or disadvantage particular groups, often inherited from biased training data or poorly chosen inputs. Learn more: AI hiring bias.
Anonymisation
Processing personal data so that individuals can no longer be identified by any reasonably likely means. Truly anonymised data generally falls outside data protection law; pseudonymised data does not. Learn more: AI and employee data privacy.
Applicant tracking system (ATS)
Software that manages job applications through the hiring process, from posting to offer. Many ATS platforms now include AI features for parsing, matching and scheduling. Learn more: ATS vs AI resume screening.
Artificial intelligence (AI)
Technology that enables computer systems to perform tasks typically associated with human intelligence, such as understanding language, recognising patterns, making predictions and generating content. Learn more: AI in HR: complete guide.
Automated employment decision tool (AEDT)
A term defined in New York City's Local Law 144 for computational tools that substantially assist or replace discretionary decisions about hiring or promotion. Employers using them must commission annual independent bias audits and notify candidates. Learn more: How to audit AI hiring tools.
Automation bias
The tendency of people to over-trust recommendations from automated systems, even when they are wrong. It can turn human review of AI decisions into a rubber stamp. Learn more: Ethical AI in HR.
B
Bias audit
An independent evaluation of an AI tool's outcomes across demographic groups to detect adverse impact. Required for automated employment decision tools in New York City. Learn more: How to audit AI hiring tools.
C
Chatbot
Software that holds a conversation with users in text or voice. Modern HR chatbots use large language models to answer employee questions from approved policy documents. Learn more: HR chatbots.
Context window
The amount of text a large language model can consider at once, including the prompt, any documents provided and the conversation so far. Content beyond the window is not taken into account. Learn more: What is an LLM.
D
Data minimisation
The data protection principle that organisations should collect and use only the personal data necessary for a specific purpose. It applies directly to the data fed into HR AI tools. Learn more: Employee data protection and AI.
Deep learning
A type of machine learning that uses neural networks with many layers to learn complex patterns. It underpins modern language, speech and image AI. Learn more: What is machine learning in HR.
Deployer
Under the EU AI Act, an organisation that uses an AI system under its authority in a professional context. An employer using a vendor's AI recruitment tool is usually a deployer and carries its own obligations. Learn more: EU AI Act employer obligations.
E
Embedding
A numerical representation of text, such as a CV or job description, that captures its meaning so AI can compare similarity. Embeddings power semantic search and candidate matching.
Explainability
The degree to which the reasons for an AI system's output can be understood and communicated to people. Essential when HR must justify decisions to candidates or employees. Learn more: AI transparency in HR decisions.
F
Fine-tuning
Further training of a pre-trained AI model on a smaller, specialised dataset so it performs better on a particular task or domain. Learn more: What is an LLM.
Foundation model
A large AI model trained on broad data that can be adapted to many tasks. Large language models such as those behind ChatGPT, Claude and Gemini are foundation models. Learn more: What is an LLM.
Four-fifths rule
A guideline from the US Uniform Guidelines on Employee Selection Procedures: a selection rate for any group below 80 percent of the rate for the highest-selected group is generally regarded as evidence of adverse impact. Learn more: AI hiring bias.
G
General-purpose AI model
An EU AI Act term for AI models that can competently perform a wide range of distinct tasks and be integrated into many downstream systems. Providers of these models have specific obligations. Learn more: EU AI Act and HR.
Generative AI
AI that creates new content, such as text, images, audio or code, in response to a prompt. In HR it drafts job descriptions, communications and learning materials. Learn more: Generative AI for HR.
Grounding
Connecting an AI model's responses to specific, trusted sources, such as an organisation's policy documents, so answers are based on verified information rather than general training data. Learn more: HR chatbots.
Guardrails
Technical and policy controls that limit what an AI system can do or say, such as blocking certain topics, restricting data access or requiring human approval. Learn more: AI policy for employees.
H
Hallucination
When a generative AI model produces information that sounds plausible but is false or fabricated, such as an invented policy clause or citation. Learn more: What is an LLM.
High-risk AI system
An EU AI Act category for AI with significant potential to affect health, safety or fundamental rights. It includes AI used in recruitment, promotion, termination, task allocation and worker monitoring. Learn more: High-risk hiring systems.
Human in the loop
A design where a person reviews, approves or can override an AI system's output before it takes effect. Meaningful human review is expected for consequential HR decisions. Learn more: Ethical AI in HR.
L
Large language model (LLM)
An AI model trained on vast amounts of text to understand and generate language by predicting likely sequences of words. LLMs power most generative AI tools used in HR. Learn more: What is an LLM.
M
Machine learning (ML)
A branch of AI in which systems learn patterns from data to make predictions or decisions, rather than following explicitly programmed rules. Learn more: What is machine learning in HR.
Model drift
The decline in an AI model's accuracy over time as real-world conditions change from those in its training data, such as shifts in the labour market or job requirements. Learn more: What is machine learning in HR.
Multimodal AI
AI that can process and generate more than one type of data, such as text, images and audio together.
N
Natural language processing (NLP)
The branch of AI that enables computers to read, interpret and generate human language. In HR it parses CVs, analyses survey comments and powers chatbots. Learn more: How AI screens resumes.
Neural network
A computing structure loosely inspired by the brain, made of layers of connected nodes that learn to recognise patterns in data. The basis of deep learning. Learn more: What is machine learning in HR.
P
People analytics
The use of data and analysis about the workforce to improve decisions on hiring, retention, performance and organisation design. Learn more: People analytics for retention.
Personal data
Any information relating to an identified or identifiable person. Most HR data, including CVs, performance records and survey responses linked to individuals, is personal data. Learn more: AI and employee data privacy.
Predictive analytics
Using historical data and statistical or machine learning models to estimate the likelihood of future outcomes, such as attrition or hiring demand. Learn more: Predictive analytics in HR examples.
Prompt
The instruction or question given to a generative AI model. The clarity and context of the prompt strongly shape the quality of the output. Learn more: ChatGPT prompts for HR.
Prompt engineering
The practice of designing prompts to get accurate, useful and consistent outputs from generative AI. Learn more: ChatGPT prompts for HR.
Prompt injection
An attack in which hidden or malicious instructions are placed in content an AI system reads, such as a CV or email, to manipulate its behaviour. Learn more: AI agents in recruitment.
Provider
Under the EU AI Act, the organisation that develops an AI system or model and places it on the market or puts it into service under its own name. Most HR technology vendors are providers. Learn more: EU AI Act and HR.
Pseudonymisation
Replacing identifying details with codes so data cannot be linked to a person without additional information kept separately. Pseudonymised data is still personal data under laws such as the GDPR. Learn more: Employee data protection and AI.
R
Retrieval-augmented generation (RAG)
A technique in which an AI system retrieves relevant passages from trusted documents and uses them to generate its answer, improving accuracy and allowing citations. Learn more: What is an LLM.
S
Sentiment analysis
Using NLP to identify the emotional tone of text, such as whether survey comments are positive, negative or neutral.
Skills taxonomy
A structured catalogue of skills and how they relate to one another and to jobs. AI uses skills taxonomies to match people to roles, learning and projects. Learn more: AI skills gap analysis.
Supervised learning
Machine learning in which a model learns from examples labelled with the correct answer, such as past employees marked as having left or stayed. Learn more: What is machine learning in HR.
T
Token
A unit of text, often a word or part of a word, that a large language model processes. Model limits and costs are usually measured in tokens. Learn more: What is an LLM.
Training data
The data used to teach an AI model. Its quality, representativeness and history directly shape the model's accuracy and fairness. Learn more: AI hiring bias.
U
Unsupervised learning
Machine learning that finds structure in unlabelled data, such as grouping employees into clusters with similar skills or survey responses. Learn more: What is machine learning in HR.
Related guides
- Common AI Terms Explained Simply: A Guide for HR Teams
Fifteen core AI terms with everyday analogies, HR examples and commonly confused pairs.
- What Is Machine Learning in HR? How It Works, Uses and Limits
How machine learning learns from people data, where HR uses it, and its limitations.
- What Is an LLM? Large Language Models Explained for HR
How large language models work, why they hallucinate, and how HR can use them safely.
- AI Literacy in the Workplace: What It Is and What the EU AI Act Requires
What AI literacy means, the amended Article 4 obligation, and a proportionate programme design.
Frequently asked questions
What AI terms should HR professionals know?
At a minimum: artificial intelligence, machine learning, generative AI, agentic AI, large language model, prompt, hallucination, training data, algorithmic bias, adverse impact, explainability and human in the loop, plus the EU AI Act terms provider, deployer and high-risk AI system.
What is the difference between AI and machine learning?
Artificial intelligence is the broad field of making computers perform tasks associated with human intelligence. Machine learning is one approach within AI, in which systems learn patterns from data. See what is machine learning in HR.
What does hallucination mean in AI?
Hallucination is when a generative AI model produces plausible-sounding information that is false or invented, such as a non-existent policy clause or citation. It is why AI outputs must be checked against sources.
What is a deployer under the EU AI Act?
A deployer is an organisation using an AI system under its authority in a professional context. Employers using AI recruitment or HR tools are usually deployers and have their own obligations, including human oversight.