Skip to content

Malicious intent in disinformation

Malicious intent refers to the deliberate motivations and strategic goals of actors who create and spread disinformation. Understanding intent is crucial for disinformation research because it distinguishes intentional falsehood from misinformation (unintentional errors) and reveals the underlying drivers of information manipulation.

Defining malicious intent

A disinformation narrative embodies malicious intent when it is deliberately designed, presented, and promoted to intentionally cause public harm or for profit (adapted from the European Commission's High Level Expert Group definition). Identifying intent requires understanding:

  • Actor goals: What outcome is the actor trying to achieve? (e.g., sowing distrust, shifting political opinion, promoting commercial products, advancing ideological narratives)
  • Narrative strategy: What false or misleading framing is deployed? How does it distort reality to achieve the goal?
  • Target audience: Who is the disinformation aimed at? What vulnerabilities or existing beliefs are being exploited?

Common malicious intent categories

Research has identified recurring patterns of disinformation intent:

Undermining the Credibility of Public Institutions (UCPI): Deliberate campaigns to weaken public trust in governmental, educational, or scientific institutions through false allegations of corruption, incompetence, or malfeasance. Example: Claims that vaccine regulators are suppressing adverse event data.

Changing Political Views (CPV): Disinformation designed to shift voter opinion or behavior ahead of elections, referenda, or policy decisions. Often targets specific constituencies with tailored narratives. Example: False claims about a candidate's funding sources.

Undermining International Organizations and Alliances (UIOA): Disinformation aimed at severing diplomatic relationships, delegitimizing multilateral bodies (UN, NATO, EU), or creating division between allied nations. Often includes false attributions of agency misconduct.

Promoting Social Stereotypes/Antagonisms (PSSA): Content designed to amplify inter-group hostility by promoting dehumanizing stereotypes, conspiracy theories targeting specific groups, or false claims about group behavior. Example: False claims about immigrant crime rates.

Promoting Anti-scientific Views (PASV): Deliberate spread of false health, environmental, or scientific information to promote alternative (often commercially-motivated or ideological) narratives. Example: Misinformation about vaccine safety or climate science.

Why intent matters

Research precision: Conflating intentional disinformation with unintentional misinformation obscures the true scope of malicious behavior. Researchers studying "fake news" datasets risk mixing distinct phenomena.

Detection and intervention: Intent classification enables more targeted countermeasures. A message promoting health misinformation may require different fact-checking responses than one designed to undermine electoral integrity.

Accountability: Understanding intent distinguishes negligent actors (media outlets repeating unchecked claims) from coordinated campaigns (state-sponsored influence operations).

Legal and policy implications: Many jurisdictions' disinformation regulations depend on demonstrating intent. The EU Code of Practice on Disinformation, for example, targets deliberate rather than inadvertent false claims.

Challenges in assessing intent

Opacity of motivation: Intent is often latent and difficult to verify. A publisher may claim good-faith error while benefiting from the disinformation they spread.

Multiple concurrent intents: A single article may combine multiple intents (e.g., undermining institutional credibility while promoting a commercial product).

Context-dependent intent: The same narrative can have different intents depending on who spread it and where. A claim may originate as state propaganda but be amplified by civilians with different motivations.

Annotator subjectivity: Assessing intent requires human judgment and inference from limited evidence, introducing annotation bias.

Key papers

  • Modzelewski et al. (2026) — MALINT: Introduces the first large-scale English corpus annotating both credibility and malicious intent. Defines a five-category intent taxonomy (UCPI, CPV, UIOA, PSSA, PASV) and demonstrates that incorporating intent analysis improves LLM-based disinformation detection by ~20% in zero-shot settings.
  • Hameleers et al. (2023) — Conceptual clarification integrating actors, intentions, and techniques; grounding intent as a defining feature of disinformation rather than misinformation.