AI PSYOPS Research Glossary

An influence activity in which humans set the objective and retain approval authority while AI supports tasks such as research, translation, drafting, analysis, or measurement.

The analytical process of linking an operation to an actor, sponsor, infrastructure, or organizational chain using multiple forms of evidence.

The use of software to execute repeated, scheduled, or predefined actions at speed or scale.

Software that pursues an influence-related objective through communication or online action with limited ongoing human direction, potentially using memory, planning, tools, and feedback.

Engagement records observable activity such as views, clicks, shares, replies, or time spent. Persuasion concerns a change in belief, attitude, decision, or behavior.

Delegating the formation, maintenance, or revision of beliefs to an AI system rather than using it only as an input to human judgment.

Influence exercised through a threat of harm or penalty, or through a severe restriction of meaningful alternatives.

A person’s capacity to form beliefs, evaluate information, and make decisions without covert subversion or unjustified interference.

An attempt to infer a person’s internal emotion, intention, stress, or vulnerability from facial, vocal, linguistic, physiological, or behavioral signals.

Covert, deceptive, exploitative, or highly asymmetric influence that subverts deliberation by exploiting cognitive, emotional, or situational vulnerabilities.

A relationship-like emotional bond with a mediated or simulated figure, including a responsive AI companion or persona.

The tailoring of content, timing, interfaces, or recommendations using a person’s stated preferences, history, context, or inferred interests.

The tailoring of messages to psychological or personality traits inferred from digital traces or profiling systems.

A relationship in which a user treats AI as a trusted interpreter of facts, values, identity, emotions, appropriate behavior, or important decisions.

A model tendency to agree with, flatter, or validate a user’s stated views and preferences even when they are inaccurate, inconsistent, or unsafe.

An industry provenance standard that uses cryptographically signed metadata to describe a media item’s origin and edit history.

A record of a media item’s origin, capture or authorship, transformations, and chain of custody or publication.

Coordinated use of deceptive accounts, identities, or assets to mislead audiences about who is acting or how popular a narrative appears.

Synthetic or manipulated audio, video, or imagery designed to create a false impression that a real or plausible person said or did something, or that an event occurred.

The advantage a wrongdoer may gain by dismissing authentic evidence as AI-generated in an environment where synthetic media are widely understood to be possible.

Text, images, audio, video, documents, or multimodal material generated or substantially transformed by artificial intelligence.

A fabricated identity presented as a real person, organization, expert, witness, activist, journalist, or community member.

Embedding a machine-detectable signal into generated text, imagery, audio, or video to indicate origin or the use of a generation system.

Intentional or structurally induced shaping of what people encounter, notice, regard as important, believe to be popular, or treat as credible through ranking, recommendation, search, moderation, or distribution systems.

A mathematical framework that adds calibrated randomness so aggregate analysis reveals limited information about any one person’s contribution.

A coordinated or emergent collection of accounts, agents, media assets, or automated systems that produces, varies, distributes, and amplifies misleading narratives at high speed or scale.

Cognitive fatigue produced by persistent informational volume, contradiction, uncertainty, and verification demands that make shared understanding harder to sustain.

The use of data analysis, machine learning, simulation, or forecasting to predict collective behavior and guide interventions intended to prevent, redirect, contain, or respond to social outcomes.

A dynamic in which acting on a prediction changes the environment and the future data used to evaluate or retrain the predictive system.

The appearance of broad, independent agreement manufactured through coordinated artificial identities, generated content, or inauthentic engagement.