Last updated: 2026-11-22
Every published row carries two independent axes: what kind of activity it is (category) and where it physically happens (domain). This page explains both, and the limits of reading counts built from them.
Category. The taxonomy has ten top-level categories, a subcategory under each, and an optional incident-type leaf. Analysts record the deepest level the evidence defends; AI suggestions are drafts until a Completer accepts them. A leaf is left empty only when its subcategory lists no leaves.
Domain. The domain is the recorded physical setting, shown on each signal and available as a filter on the signals list:
Unknown means the domain was not recorded; it is not a claim that the activity has no physical setting. The method for both axes is documented under physical domain in the methodology.
These are the ten categories. The subcategories below each name are the recorded second level; the incident-type leaf is the third level. The list is not a ranking of severity. The count beside each category is the number of published records in it, from the same visibility rule as the signals list.
Military & Paramilitary Operations2576
Conventional Military Pressure · Unconventional / Paramilitary · High-Tech Military Tools
Cyber & Information Warfare862
Cyber Attacks · Information Operations · Psychological Warfare
Political & Legal Subversion454
Political Manipulation · Lawfare & Legal Pressure · Diplomatic Tools
Intelligence & Espionage631
Human Intelligence (HUMINT) · Signals/Technical Intelligence (SIGINT/TECHINT) · Counterintelligence Sabotage
Energy & Infrastructure Attacks234
Critical Infrastructure Sabotage · Resource Weaponization
Societal & Cultural Pressure80
Vandalism / Symbolic Attacks · Propaganda & Cultural Manipulation
Population & Migration Tools196
Weaponized Migration · Human Trafficking · Demographic Pressure
Terrorism & Criminal Proxy Tactics295
Terrorism · Organized Crime
Economic & Financial Pressure152
Sanctions Evasion & Smuggling · Financial Influence · Corruption as a Tool
Environmental / Non-traditional Tools23
Environmental Sabotage · Bio/Chemical Threats · Climate/Resource Manipulation
A row is listed once in the signals list, but a single occurrence can affect several countries. On country views, that row appears in each affected country, so country totals may exceed the number of distinct incidents. The same applies when one occurrence spans several domains.
Category counts sum to rows in the current view, not to distinct occurrences in the world. Compare trends over time with the insights page rather than adding totals from different filters.
HWT tracks hybrid-threat activity that affects EU/EFTA/EEA and NATO Europe. Pure Ukraine-theater kinetic reporting without a Europe/NATO hybrid effect, routine domestic crime without a hybrid vector, and unverified social posts without a corroborating source path are out of scope. The full inclusion bar is under inclusion in the methodology.
The category describes the activity, not who carried it out. Attribution is a separate, derived presentation: a reliability band (Strong or Pattern fit), an attribution-confidence band, and an attribution basis (single source or independence confirmed). None of these is a third taxonomy level, and a confident category does not imply a confident attribution.