AI literacy’ means skills, knowledge and understanding that allow providers, deployers and affected persons, taking into account their respective rights and obligations in the context of this Regulation, to make an informed deployment of AI systems, as well as to gain awareness about the opportunities and risks of AI and possible harm it can cause.
Source: AI Act, Article 3(56)
A machine-based system that is designed to operate with varying levels of autonomy and may exhibit adaptiveness after deployment. It uses input data to produce outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments.
Source: AI Act, Article 3(1)
A structured set of rules or instructions designed to solve a problem or perform a task. In AI, algorithms guide how machines process input data to produce an output, with the intention of mimicking human intelligence.
Source: JRC , Glossary of human-centric artificial intelligence, 2022.
An inclination of prejudice towards or against certain groups of people, objects or positions in the outputs of an AI system, often resulting in discriminatory outcomes. Bias can arise in different ways, such as through training data (for example, when data about some groups are underrepresented), the design of rules and choices made by developers, ongoing learning and adaptation through interaction, or personalisation features.
Source: HLEG AI, Ethics Guidelines for Trustworthy AI, 2019.
Hypothetical ability of an intelligent system that can successfully understand, learn and perform any intellectual task that a human being can. Also referred to as strong AI.
Source: JRC, Glossary of human-centric artificial intelligence, 2022.
A set of sciences, theories and techniques whose purpose is to reproduce by a machine the cognitive abilities of a human being.
Source: Council of Europe, Artificial Intelligence Glossary.
Automated decision-making can refer to both solely automated decisions and automated assisted decision-making. Solely automated decision-making means decisions that are fully automated with no human judgement. Automated assisted decision-making is when automated or algorithmic systems assist human judgement and decision-making.
Source: UK Parliament, Artificial intelligence (AI) glossary, 2024.
An all-encompassing term for any collection of data sets so large or complex that they are difficult to store, manage and process with conventional, non-scalable technology.
Source: JRC, Glossary of human-centric artificial intelligence, 2022.
Personal data resulting from specific technical processing relating to the physical, physiological or behavioural characteristics of a natural persona, which allow or confirm the unique identification of that natural person, such as facial images or dactyloscopic data.
Source: General Data Protection Regulation (GDPR), Article 4(14)
Automated recognition of physical, physiological and behavioural human features such as the face, eye movement, body shape, voice, prosody, gait, posture, heart rate, blood pressure, odour, keystrokes characteristics, for the purpose of establishing an individual’s identity by comparing biometric data of that individual to stored biometric data of individuals in a reference database.
Source: AI Act, Article 3 (35), Recital (15)
A system which can be viewed in terms of its inputs and outputs, without any knowledge of its internal workings.
Source: JRC, Glossary of human-centric artificial intelligence, 2022.
A computer programme designed to simulate conversation with a human user, usually over the internet; especially one used to provide information or assistance to the user as part of an automated service.
Source: Oxford English Dictionary, Chatbot.
A field of AI focused on programming computer systems to interpret and understand images, videos and other visual inputs and take actions or make recommendations based on that information. Applications include object recognition, facial recognition, medical imaging analysis, navigation and video surveillance.
Source: UK Parliament, Artificial intelligence (AI) glossary, 2024.
The practice of protecting computers, servers, networks, data and systems from malicious attacks, unauthorised access and damage.
Source: Council of the EU and the European Council, Cybersecurity.
Computational process that extracts patterns by analysing quantitative data from different perspectives and dimensions, categorising it, and summarising potential relationships and impacts.
Source: ISO/IEC DIS 22989(en), Terms related to Artificial Intelligence.
A subset of machine learning that uses artificial neural networks to recognise patterns in data and provide a suitable output, for example, a prediction. Deep learning is suitable for complex learning tasks, and has improved AI capabilities in tasks such as voice and image recognition, object detection and autonomous driving.
Source: UK Parliament, Artificial intelligence (AI) glossary, 2024.
AI-generated or manipulated image, audio or video content that resembles existing persons, objects, places, entities or events and would falsely appear to a person to be authentic or truthful.
Source: AI Act, Article 3(60)
A natural or legal person, public authority, agency or other body using an AI system under its authority except where the AI system is used in the course of a personal non-professional activity.
Source: AI Act, Article 3(4)
An AI system for the purpose of identifying or inferring emotions or intentions of natural persons on the basis of their biometric data.
Source: AI Act, Article 3(39)
Term used to indicate the development, deployment and use of AI that ensures compliance with ethical norms, including fundamental rights, ethical principles, and related core values.
Source: HLEG AI, Ethics Guidelines for Trustworthy AI, 2019.
Property of an AI system to express important factors influencing the AI system results in a way that humans can understand.
Source: ISO/IEC DIS 22989(en), Terms related to Artificial Intelligence.
Automatic processing of digital images containing individuals’ faces for identification or verification of those individuals by using face templates.
Source: Council of Europe, Consultative Committee of the Convention for the protection of individuals with regard to automatic processing of personal data (Convention 108), Guidelines on facial recognition, 2021.
Large AI models trained on a vast quantity of data that can be adapted to a wide range of downstream tasks. Also referred to as “foundation(al) models”.
Source: JRC, Glossary of human-centric artificial intelligence, 2022.
An AI model that generates text, images, audio, video or other media in response to user prompts. It uses machine learning techniques to create new data that has similar characteristics to the data it was trained on. Generative AI applications include chatbots, photo and video filters, and virtual assistants.
Source: UK Parliament, Artificial intelligence (AI) glossary, 2024.
Plausible sounding but inaccurate outputs generated by large language models (such as ChatGPT).
Source: UK Parliament, Artificial intelligence (AI) glossary, 2024.
An AI system that can significantly affect people’s health, safety, or fundamental rights, such as AI used in biometric identification, education, employment, essential services, law enforcement, or justice, and is subject to safeguards under the AI Act.
Source: AI Act, Article 6(2) and Annex III
The human-centric approach to AI strives to ensure that human values are central to the way in which AI systems are developed, deployed, used and monitored, by ensuring respect for fundamental rights, including those set out in the EU Charter of Fundamental Rights. This also entails consideration of the natural environment and of other living beings that are part of the human ecosystem, as well as a sustainable approach enabling the flourishing of future generations to come.
Source: HLEG AI, Ethics Guidelines for Trustworthy AI, 2019.
The capability for human intervention in every decision cycle of the system (human-in-the-loop), or during the design cycle of the system and monitoring the system’s operation (human-on-the-loop), or the capability to oversee the overall activity of the system and the ability to decide when and how to use the system in any particular situation (human-in-command). Human oversight can include the decision not to use the system in a particular situation, to establish levels of human discretion during the use of the system, or to ensure the ability to override a decision made by a system.
Source: JRC, Glossary of human-centric artificial intelligence, 2022 (based on HLEG AI, Assessment List for Trustworthy AI (ALTAI))
Interpretability refers to the concept of comprehensibility, explainability, or understandability. When an element of an AI system is interpretable, this means that it is possible at least for an external observer to understand it and find its meaning.
Source: HLEG AI, Assessment List for Trustworthy AI (ALTAI)
This is a privacy-preserving technique for data anonymisation. K-anonymity ensures that each individual in a dataset cannot be distinguished from at least k-1 other individuals with respect to the quasi-identifiers in the dataset. This means the data is not unique to a certain individual and makes it more difficult for an attacker to re-identify specific individuals in the dataset.
Source: University of Utrecht, Data Privacy Handbook: K-anonymity, l-diversity and t-closeness.
A type of foundation model (General Purpose AI model) that is trained on vast amounts of text to carry out natural language processing tasks. During training phases, large language models learn parameters from factors such as the model size and training datasets. Parameters are then used by large language models to infer new content.
Source: UK Parliament, Artificial intelligence (AI) glossary, 2024.
A branch of AI and computer science which focuses on development of systems that are able to learn and adapt without following explicit instructions imitating the way that humans learn, gradually improving its accuracy, by using algorithms and statistical models to analyse and draw inferences from patterns in data.
Source: JRC, Glossary of human-centric artificial intelligence, 2022
Term used to describe AI systems that are specified to handle a singular or limited task. Many currently existing AI systems are operating as a narrow AI focused on a specific problem. For instance, digital assistants are all examples of narrow AI as they operate within a limited pre-defined range of functions.
Source: JRC, Glossary of human-centric artificial intelligence, 2022
A field of AI focused on programming computer systems to understand and generate human speech and text. Algorithms look for linguistic patterns in how sentences and paragraphs are constructed and how words, context and structure work together to create meaning. Applications include speech-to-text converters, online tools that summarise text, chatbots, speech recognition and translations.
Source: UK Parliament, Artificial intelligence (AI) glossary, 2024.
Underlying code used to run AI models is freely available for testing, scrutiny and improvement.
Source: UK Parliament, Artificial intelligence (AI) glossary, 2024.
Any information relating to an identified or identifiable natural person (‘data subject’); an identifiable natural person is one who can be identified, directly or indirectly, in particular by reference to an identifier such as a name, an identification number, location data, an online identifier or to one or more factors specific to the physical, physiological, genetic, mental, economic, cultural or social identity of that natural person.
Source: General Data Protection Regulation (GDPR), Article 4(1)
A breach of security leading to the accidental or unlawful destruction, loss, alteration, unauthorised disclosure of, or access to, personal data transmitted, stored or otherwise processed. Source: General Data Protection Regulation (GDPR), Article 4(12)
Any operation or set of operations performed or not using automated processes and applied to personal data or sets of data, such as collection, recording, organisation, structuring, storage, adaptation or modification, retrieval, consultation, use, communication by transmission, dissemination or any other form of making available, linking or interconnection, limitation, erasure or destruction.
Source: Council of Europe, Artificial Intelligence, Glossary.
A natural or legal person, public authority, agency or other body that develops an AI system or a general-purpose AI model or that has an AI system or a general-purpose AI model developed and places it on the market or puts the AI system into service under its own name or trademark, whether for payment or free of charge.
Source: AI Act, Article 3(3)
Processing of personal data in such a manner that the personal data can no longer be attributed to a specific data subject without the use of additional information, provided that such additional information is kept separately and is subject to technical and organisational measures to ensure that the personal data are not attributed to an identified or identifiable natural person.
Source: General Data Protection Regulation (GDPR), Article 4(5)
Refers to the practice of designing, developing, and deploying AI with certain values, such as being trustworthy, ethical, transparent, explainable, fair, robust and upholding fundamental rights.
Source: UK Parliament, Artificial intelligence (AI) glossary, 2024.
Machines that are capable of automatically carrying out a series of actions and moving in the physical world. Modern robots contain algorithms that typically, but do not always, have some form of artificial intelligence. Applications include industrial robots used in manufacturing, medical robots for performing surgery, and self-navigating drones.
Source: UK Parliament, Artificial intelligence (AI) glossary, 2024.
Ability of a system to maintain its level of performance under any circumstances, including the ability to deal with execution errors, erroneous inputs, or unseen data. Robustness of an AI system encompasses both its technical robustness, as well as its robustness from a social perspective (ensuring that the AI system duly takes into account the context and environment in which the system operates).
Source: JRC, Glossary of human-centric artificial intelligence, 2022
An incident or malfunctioning of an AI system that directly or indirectly leads to any of the following:
(a) the death of a person, or serious harm to a person’s health;
(b) a serious and irreversible disruption of the management or operation of critical infrastructure.
(c) the infringement of obligations under Union law intended to protect fundamental rights;
(d) serious harm to property or the environment;
Source: AI Act, Article 3(49)
Personal data revealing racial or ethnic origin, political opinions, religious or philosophical beliefs, or trade union membership, and the processing of genetic data, biometric data for the purpose of uniquely identifying a natural person, data concerning health or data concerning a natural person’s sex life or sexual orientation
Source: General Data Protection Regulation (GDPR), Article 9(1)
A way of training machine learning systems for a specific application. In a training phase, an AI system is fed labelled data. The system trains from the input data, and the resulting model is then tested to see if it can correctly apply labels to new unlabelled data (such as if it can correctly label unlabelled pictures of cats and dogs accordingly).
Source: UK Parliament, Artificial intelligence (AI) glossary, 2024.
This is information that has been generated on a computer using statistical methods of AI techniques. Synthetic data mimics real data and can be used to augment or replace real data in the training of AI models to improve their performance, protect sensitive data, and mitigate bias.
Source: IBM, What is synthetic data?
Data used for providing an independent evaluation of the AI system in order to confirm the expected performance of that system before its placing on the market or putting into service.
Source: AI Act, Article 3(32)
Data used for training an AI system through fitting its learnable parameters. Training datasets can be labelled (for example, pictures of cats and dogs labelled ‘cat’ or ‘dog’ accordingly) or unlabelled.
Source: AI Act, Article 3(29); UK Parliament, Artificial intelligence (AI) glossary, 2024.
Trustworthy AI has three components: (1) it should be lawful, ensuring compliance with all applicable laws and regulations (2) it should be ethical, demonstrating respect for, and ensure adherence to, ethical principles and values and (3) it should be robust, both from a technical and social perspective, since, even with good intentions, AI systems can cause unintentional harm.
Source: HLEG AI, Assessment List for Trustworthy AI (ALTAI)
A way of training machine learning systems for a specific application. An AI system is fed large amounts of unlabelled data, in which it starts to recognise patterns of its own accord. This type of learning is useful when it is not clear what patterns are hidden in data, such as in online shopping basket recommendations.
Source: UK Parliament, Artificial intelligence (AI) glossary, 2024.
Data used for providing an evaluation of the trained AI system and for tuning its non-learnable parameters and its learning process in order, inter alia, to prevent underfitting or overfitting; whereas the validation dataset can be a separate dataset or part of the training dataset, either fixed or variable split.
Source: AI Act, Article 3(31)