Our artificial intelligence (AI) glossary for HR teams and managers covers over 60 important AI definitions. From AI productivity to AI workplace safety, our glossary includes terms that will help to improve your understanding of this important new technology, with a focus on how AI is changing the workplace and its implications for employee mental health and wellbeing.
A to Z AI Glossary
AI Champions Network
A network of employees who support colleagues in understanding and adopting AI, sharing experiences, identifying opportunities and promoting responsible use.
AI Change Management
The process of supporting employees and organisations through the introduction of AI, including changes to working practices, responsibilities, skills and organisational culture.
AI Explainability
The extent to which an AI system's outputs or decisions can be understood and explained, particularly when AI is used to support recruitment, performance management or other employment decisions.
AI-Generated Grievances
The use of generative AI by employees to draft grievances. AI can help people put genuine concerns into words, but can also produce lengthy or complex submissions that make the underlying issue harder to identify and resolve. Similar patterns are also showing up with subject access requests and tribunal claims.
AI in Recruitment
The use of artificial intelligence to support recruitment processes, including writing job descriptions, screening applications, assessing candidates and identifying suitable applicants, while raising considerations around fairness, transparency and human oversight.
AI Job Insecurity (sometimes called Existential Anxiety)
Concerns about the potential impact of AI on an individual's employment, career prospects, professional relevance or long-term job security. It is one of the recognised forms of technostress.
AI Literacy
The knowledge and skills needed to understand, use and critically evaluate AI tools effectively, including recognising their capabilities, limitations and potential risks.
AI Policy
An organisation's formal guidelines outlining how AI tools can be used, including approved applications, employee responsibilities, data protection requirements and appropriate safeguards. Take a look at Clay’s AI Writing Policy.
AI Productivity
Improvements in the efficiency, quality or quantity of work achieved through AI, including automating routine tasks, supporting decision-making and enhancing employee capabilities.
AI Safety
The practices, safeguards and processes designed to prevent or minimise harm arising from AI systems, including risks to individuals, organisations and wider society.
AI Self-Efficacy
An individual’s belief in their ability to learn, use and adapt to AI tools effectively. It is distinct from how well the AI system itself performs.
AI System Effectiveness (sometimes called AI Efficacy)
The extent to which an AI system achieves its intended outcomes or performs its designated tasks effectively.
AI Tools
Software applications or systems that use artificial intelligence to perform or support tasks such as analysing information, generating content, automating processes and assisting decision-making.
AI TRiSM (AI Trust, Risk and Security Management)
A framework popularised by Gartner for managing AI-related risks, including security, privacy, fairness, reliability and the monitoring of AI systems.
Algorithmic Bias
Systematic errors or unfair outcomes produced by algorithms, which can arise from biased data, system design or implementation and may disadvantage particular individuals or groups. Where algorithmic bias leads to less favourable outcomes for people with a protected characteristic, it may amount to discrimination under the Equality Act 2010. The employer remains responsible for the outcome even when the tool is supplied by a third party.
Algorithmic Management
The use of algorithms or AI to allocate, schedule, monitor or evaluate work and workers, taking on some tasks traditionally carried out by line managers. It raises questions about transparency, fairness, autonomy and employee wellbeing.
Artificial Intelligence (AI)
Technology that enables computer systems to perform tasks typically associated with human intelligence, including recognising patterns, analysing information, understanding language and generating content.
Automated Decision-Making (UK GDPR)
Rules governing decisions made solely by automated means that have legal or similarly significant effects on individuals, such as recruitment or dismissal decisions. Since February 2026, UK organisations making such decisions must tell individuals, allow them to make representations, provide meaningful human intervention and allow them to contest the decision. Stricter limits apply where special category data is used, and organisations dealing with individuals in the EU may still be subject to the original EU GDPR Article 22.
Automated Surveillance
The use of technology to automatically monitor, record or analyse employee activities, such as productivity, computer usage or location, which can raise concerns about privacy, trust and wellbeing.
Automation Bias
The tendency to give excessive weight to automated recommendations or outputs, sometimes accepting them without sufficient independent checking. In AI-enabled work, appropriate review and human judgement remain important.
Brain Fry
An informal term describing mental exhaustion associated with intensive cognitive demands, including repeatedly reviewing, checking or evaluating AI-generated information.
Calibrated Trust in AI
Trust that matches how reliable an AI system actually is for a given task and context. Well-calibrated trust means relying on AI where it performs well and checking or overriding it where it does not. Over-trust can lead to errors being accepted (see Automation Bias), while under-trust can mean useful tools are ignored.
Change Fatigue
The physical, mental or emotional exhaustion employees may experience when adapting to frequent or ongoing organisational changes, including the introduction of new technologies.
Chatbot
A software application designed to interact with users through written or spoken conversation, often using AI to answer questions, provide information or assist with tasks.
Cognitive Offloading
The practice of using external tools, including AI, to reduce the mental effort required for tasks such as remembering information, solving problems or processing information.
Cognitive Overload
A state in which the amount or complexity of information and mental demands exceeds an individual's capacity to process it effectively, potentially affecting concentration, decision-making and performance.
Custom AI Assistants / Agents
AI systems configured for specific organisational needs, such as answering employee questions, supporting HR administration or carrying out defined tasks. Some agents can independently perform actions within established boundaries.
Data Handling
The processes and safeguards used to collect, store, access, share and process information, including ensuring employee and organisational data is managed appropriately when using AI.
Decision-Making (AI-Assisted)
The use of AI-generated information, predictions or recommendations to support decisions made by individuals or organisations, including decisions relating to recruitment, performance and workforce planning.
Digital Transformation
The integration of digital technologies into organisational processes, services and working practices, often involving changes to roles, skills, operations and organisational culture.
"Do Not Share" Data Guardrails
Organisational rules identifying information that employees must not enter into AI tools, particularly confidential business information, personal data and sensitive employee records.
Employee Experience Personalisation
The use of technology, including AI, to tailor aspects of the employee experience, such as onboarding, learning opportunities, communication and career development, to individual needs and preferences.
Employee Voice
The opportunities and mechanisms through which employees express their views, share concerns and influence workplace decisions, including those relating to the introduction and use of AI.
Ethical AI
The development and use of artificial intelligence in ways that consider fairness, transparency, accountability, privacy, human rights and the potential consequences for individuals and society.
EU AI Act
European Union legislation that regulates AI according to its level of risk. AI used in recruitment, promotion, task allocation and performance monitoring is classed as high-risk, bringing obligations around human oversight, transparency and data quality. The EU Artificial Intelligence Act may apply to UK organisations where their AI systems are used in relation to people in the EU.
Generative AI
Artificial intelligence capable of creating new content, including text, images, audio, video and computer code, based on patterns learned from data and instructions provided by users.
Generative Engine Optimisation (GEO)
Like Search Engine Optimisation (SEO), Generative Engine Optimisation is the process of improving the visibility of your website and brand in responses generated by AI tools.
Governance (AI Governance)
The policies, processes, responsibilities and oversight arrangements used to manage how AI is developed, introduced and used within an organisation.
Grounded Knowledge Base
A collection of trusted information, such as internal policies and organisational documents, that an AI system can use to inform its responses, helping improve relevance and accuracy without eliminating the possibility of errors.
Hallucination
When an AI system produces information that sounds plausible but is inaccurate, fabricated or unsupported, such as invented facts, references or policy details. It is one of the main reasons human review of AI outputs remains essential.
Human Agency and Autonomy
The extent to which employees retain meaningful choice, influence and control over how their work is carried out and how AI is used.
Human-AI Collaboration
The process of people and AI systems working together to complete tasks, combining human judgement, experience and interpersonal skills with AI capabilities such as data analysis, automation and content generation. Better outcomes are not guaranteed: research suggests that, on average, human–AI combinations perform worse than the stronger of the two working alone, so how the collaboration is designed matters.
Human-Centred AI
An approach to designing, implementing and using AI that prioritises human needs, rights, capabilities and wellbeing alongside organisational and technological objectives.
Human-in-the-Loop (HITL)
An approach to AI in which human involvement is built into a process, allowing individuals to review, guide, approve or correct AI outputs or decisions at defined stages.
Human Oversight
The involvement of people in monitoring, reviewing and, where appropriate, intervening in AI-supported processes to help ensure outputs and decisions are accurate, appropriate and accountable.
Job Demands–Resources (JD–R) Model
A framework used to understand how workplace demands, such as workload and pressure, and resources, such as autonomy and support, influence employee wellbeing, motivation and performance. It can also be applied to changes introduced by AI.
Job Displacement
The loss or reduction of employment opportunities when technology, including AI, replaces or significantly reduces the need for particular roles or types of work.
Job Redesign
The process of changing the tasks, responsibilities or structure of a role to reflect evolving organisational needs, including opportunities created by AI and automation.
Large Language Model (LLM)
A type of AI model trained on large amounts of data, particularly text, to process and generate language, enabling applications such as AI assistants, content creation and information summarisation.
Natural Language Processing (NLP)
A field of artificial intelligence concerned with enabling computers to analyse, interpret and generate human language, including written and spoken communication.
Predictive Health / Attrition Analytics
The use of statistical models and data analysis, potentially including AI, to identify patterns associated with employee absence, wellbeing or the likelihood of employees leaving an organisation. Such analysis requires appropriate safeguards for privacy, fairness and data protection.
Professional Identity Anxiety (sometimes called Identity Threat)
Concerns about the impact of AI on an individual's professional identity, sense of purpose or perceived value, particularly when technology takes over tasks previously central to their role.
Prompt Engineering
The practice of designing and refining instructions given to AI systems to improve the relevance, accuracy and usefulness of their responses.
Prompt Guidance
Organisational advice and instructions that help employees use AI prompts effectively, including providing appropriate context, checking outputs and protecting confidential information.
Psychological Safety
A shared belief that individuals can speak up, ask questions, acknowledge mistakes and express concerns without fear of interpersonal embarrassment or punishment, including when discussing AI-related changes. Find out more about Psychological Safety.
Psychosocial Risk
The potential for harm to employees' psychological or physical health arising from how work is designed, organised or managed, including risks associated with AI-enabled monitoring, increased workloads or reduced autonomy.
Responsible AI
The development, deployment and use of AI in ways that promote fairness, transparency, accountability, safety and respect for individuals' rights and wellbeing.
Shadow AI
The use of AI tools or systems within an organisation without appropriate authorisation, visibility or oversight, potentially creating risks relating to data protection, security and compliance.
Skills Erosion
The gradual reduction of knowledge, abilities or expertise through limited use or practice, potentially occurring when employees become overly reliant on AI to perform certain tasks.
Skills Gap
The difference between the skills employees currently possess and those required to meet existing or future organisational needs, including skills associated with AI adoption.
Task Unbundling
The process of breaking down jobs into individual tasks to understand how work is performed and identify activities that could be automated, supported or redesigned using AI.
Technostress
Stress or psychological strain associated with using, adapting to or managing technology, including pressures relating to complexity, constant connectivity, changing skills requirements or digital monitoring. Common forms include concern about keeping pace with new tools and changing expectations (sometimes called operational anxiety) and worry about job security (see AI Job Insecurity).
Transactional Workplace Risk
The potential for workplace relationships and interactions to become increasingly impersonal or task-focused when automation reduces opportunities for human connection, communication and collaboration.
Trust in AI
The extent to which individuals believe AI systems are reliable, appropriate, fair and capable of performing their intended functions, influencing their willingness to use or rely on them.
TUPE & AI (Transfer of Undertakings)
The potential implications of AI-driven restructuring, outsourcing or service changes for the Transfer of Undertakings (Protection of Employment) Regulations. Whether TUPE applies depends on the specific circumstances and legal requirements of the transfer.
Work Creep
The gradual expansion of employees' tasks, responsibilities or working hours beyond their original expectations, potentially occurring when AI-enabled efficiencies lead to additional work being allocated. A related effect is work intensification, where the pace or density of work increases because time saved through AI is filled with higher output expectations rather than higher-value work, learning or recovery.
Workforce Planning
The process of assessing current and future workforce requirements and developing strategies to ensure an organisation has the people, skills and capabilities needed to achieve its objectives, including adapting to technological change.
Zero-Shot Prompting
Asking a generative AI tool to complete a task when you don’t provide an example beforehand. E.g., asking a generative AI tool to draft a wellbeing policy without uploading an example policy to base it on.
Spotted a relevant term that we’ve missed in our AI glossary? Please let us know and we’ll add it in.
How Altruist can help
Are you planning to implement AI tools within your workforce, or do you already use AI but aren’t sure how it's affecting your employees?
According to Yulife’s AI in HR Report 2026, 45% of workers feel more pressure since generative AI tools arrived. As experts in tailored mental health and wellbeing training and consultancy, we’re here to help HR teams and managers navigate AI adoption, boosting workplace engagement and productivity while protecting their teams’ happiness and health.
If you have any questions about the people impact of AI adoption, we’d be happy to answer them.
Simply get in touch to talk with us and learn more.
