A

AI Agent Governance is the process of creating rules, systems, and oversight mechanisms to ensure that AI agents operate safely, ethically, and in alignment with human values and objectives.

Adversarial testing, or 'red teaming,' identifies risks in AI by attacking models, infrastructure, and deployment, revealing vulnerabilities for potential harm.

The degree to which an AI system’s mechanisms are open and understandable to users and stakeholders.

The principle that algorithmic decisions should not create unjust or prejudicial outcomes for certain groups of people, particularly those defined by sensitive characteristics like race, gender, or age.

The principle that entities (individuals or organizations) are responsible for how their AI systems operate and the outcomes they produce.

The ability of various stakeholders, including non-technical ones, to understand AI concepts, potential, and limitations.

A sequence of instructions or set of rules designes to complete a task or solve a problem.

The process of translating AI Ethics into practice to identify and mitigate the risks of AI systems.

A subdiscipline of the broader umbrella of digital ethics that investigates the psychological, social, and political impact of AI.

The research and practice of assessing, mitigating, and assuring an algorithm’s safety, legality, and ethics.

Any computational process, derived from machine learning, statistical modelling, data analytics, or artificial intelligence, that issues simplified output, including a score, classification, or recommendation.

A computational process, including one derived from machine learning, statistics, or other data processing or artificial intelligence techniques, that screens, evaluates, categorises, recommends, or makes or facilitates employment-related decisions.

The process of declaring that a system conforms to predetermined standards, practices or regulations.

An AI system is a machine-based system that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments.

A substantially different rate of selection in hiring, promotion, or other employment decision which works to the disadvantage of members of a race, sex, or ethnic group.

A metric for evaluating a model as being correct or precise. Accuracy is used to evaluate how well an AI model predicts or that it classifies data correctly.