Reference

Narrative Threat Glossary

Clear, plain-language definitions of the key terms in narrative threat intelligence and coordinated narrative manipulation. Each term links to a fuller explanation.

Narrative Attack: A coordinated campaign that uses information, true, false or misleading, to shape what people believe about an organisation and cause measurable harm to its value. The weapon is the coordination, not the individual post.
Narrative Risk: The business exposure an organisation carries when a story about it, true or false, spreads and damages its value: share price, reputation, customer trust or regulatory standing. A narrative attack is the event; narrative risk is the standing exposure to it.
Coordinated Inauthentic Behaviour (CIB): Activity by a group of accounts working together to mislead, while disguising the fact that they are coordinated or not genuine. CIB is the behavioural signature that behavioural detection targets.
Behavioural Threat Intelligence: The discipline of detecting threats by analysing how accounts behave, such as timing and coordination, rather than what they say. It is language-agnostic and resistant to content manipulation.
Astroturfing: Coordinated activity disguised to look like spontaneous, grassroots public opinion, used to manufacture the appearance of consensus.
Bot Network: A coordinated group of automated or semi-automated accounts controlled by a single operator and used to amplify a message far beyond what any individual could achieve.
Pump-and-Dump / Ramp-and-Dump: Market manipulation in which coordinated actors inflate an asset price with hype, then sell at the peak, leaving other investors with losses. A narrative attack pointed at a share price.
Disinformation: False information spread deliberately to deceive. Disinformation is about the content; a narrative attack is about the coordination, and can be built on true information that is artificially amplified.
Deepfake: Synthetic or manipulated media, audio, video or an image, generated or altered by AI to appear authentic and used deceptively, most often to impersonate a person or seed a false narrative.
Synthetic Media: The umbrella term for any AI-generated or AI-altered content. A deepfake is synthetic media used deceptively; not all synthetic media is a deepfake.
Content Authenticity: Whether a piece of media is genuine or has been AI-generated or manipulated. Content-authenticity detection scores media for signs of synthetic manipulation, and does not depend on provenance metadata or watermarks.
Voice Cloning: Using AI to reproduce a specific person's voice from a short audio sample, enabling synthetic speech in that voice. A common component of audio deepfakes and impersonation fraud.
Cross-Platform Correlation: Detecting the same coordinated narrative pushed across multiple platforms in the same window. Single-platform monitoring misses campaigns that are deliberately split across channels.
Deterministic and Explainable Verdict: A detection result that traces back to specific behavioural evidence, rather than a black-box score, so that risk, legal and compliance teams can act on and defend the decision.
Bot-to-Human Spectrum: A scoring model that places an account on a continuum from clearly automated to clearly human, rather than a simple yes-or-no bot label, reflecting how real inauthentic activity actually behaves.
Shared Profile Layer: A data architecture in which every account analysed enriches a shared intelligence layer for all customers, so detection accuracy compounds as usage grows.
Calibrated Confidence: Pairing every detection verdict with an honest measure of how much evidence supports it, and abstaining when that evidence is too thin, rather than returning a confident-looking score regardless.