Our standards for inclusion, sourcing, and threat tier classification.
Entities are included in this database when documented evidence demonstrates actions that threaten liberal democratic institutions, norms, or processes in the United States. We focus on:
Documented Actions
We document actions, not beliefs. Political opinions, however extreme, are not sufficient for inclusion without corresponding anti-democratic conduct.
Verifiable Evidence
All claims must be supported by credible sources. Rumors, anonymous allegations, and unverified reports are not sufficient.
Democratic Impact
The documented actions must have a clear connection to undermining democratic institutions, elections, rule of law, or constitutional governance.
Private Figures and Minors
Private figures are included only where conduct is established by conviction, guilty plea, or official designation. Named minors are excluded absent adult criminal adjudication.
What Counts as an Organization
“Organization” includes formal groups, online platforms and forums, and decentralized networks. Operations run by a single identified person are documented as an organization when the operation itself is the threat vector, with the operator as a connected individual entity.
We prioritize authoritative sources with strong editorial standards and verification processes:
Legal Documents
Court filings, indictments, verdicts, DOJ press releases, and official legal proceedings. These are our gold standard.
Government Sources
Congressional reports, official government investigations, regulatory filings, and statements from government agencies.
Academic Research
Peer-reviewed journals, university research centers, and academic experts in relevant fields.
Major Journalism
Established news organizations with robust editorial standards: AP, Reuters, major newspapers, investigative journalism outlets like ProPublica.
Entities are classified into five tiers based solely on the severity of their documented actions. The strength of the evidence behind a classification is expressed separately as an analytic confidence level (see Analytic Confidence below), never in the tier itself. Lower tier numbers indicate more severe documented conduct. Tiers describe documented past conduct, not predictions of future behavior.
Direct participation in, material support for, or operational coordination of political violence or terrorism. Applies to individuals and organizations.
Sustained rhetoric or content production that identifies targets and solicits, valorizes, or instructs violence — the pattern scholars call stochastic terrorism. We document mobilization patterns, not individual causation; direct causal attribution to specific attacks is noted where established but not required.
Concrete actions undermining electoral integrity, rule of law, or peaceful transfer of power.
Materially advancing authoritarian consolidation through policy, legislation, litigation, funding, or institutional pressure — whether as officeholder, candidate, organization, or major funder. Note: involves subjective judgment.
Individuals and organizations producing or amplifying content that legitimizes anti-democratic action or dehumanizes targeted groups, without documented Tier 1-3 conduct. Note: explicitly acknowledges subjective assessment.
Each tier has a minimum evidentiary basis. Evidence above the floor raises the analytic confidence level; it never changes the tier.
Categories describe the mechanisms behind an entity's documented conduct. They are organized into seven groups across three facets: milieu (the documented movement or ideological identity), conduct (what the entity does), and target(who documented hostility is aimed at). Categories describe documented milieus and conduct; empty categories are expected — the vocabulary's existence is the objectivity guarantee, not its population.
Designated Terrorist Organization requires formal designation, proscription, or listing by a named authority (US FTO/SDGT, UK, EU, Canada, Australia, NZ), with the designation instrument cited as a source. Individuals qualify only via personal designation (e.g., SDGT). Descriptive government vocabulary is not designation.
Nihilistic Violent Extremism follows operational FBI/DOJ usage while the term remains academically contested; we apply it only where federal filings or listings use it, and note the scholarly debate.
Great Replacement Theory is assigned only when the narrative is the documented mobilizing frame of the conduct, not merely co-occurring rhetoric.
Targeted-hate categories (antisemitism, anti-LGBTQ, anti-immigrant) name victim classes where hostility toward that class is itself the documented pattern; other victim characteristics are documented at the claim level.
Lawful-but-contested conduct categories (election denial, voter suppression, weaponized litigation, dark money, and similar) denote systematic campaigns with documented democratic impact, never isolated policy advocacy or ordinary political speech.
These requirements are enforced in software at publish time and on every subsequent edit; a label cannot exist on a published entity without disclosed supporting claims.
We assign confidence levels to our assessments following intelligence community standards. Confidence reflects the quality and corroboration of available evidence, not the certainty that an entity poses a threat.
(85-100%)
Strong, corroborated evidence from multiple reliable sources. The assessment is well-supported and unlikely to change significantly.
(60-84%)
Good evidence with some gaps or limited corroboration. Additional evidence could strengthen or modify the assessment.
(30-59%)
Limited evidence with significant uncertainty. The assessment is provisional pending additional research.
We rate source reliability using a standardized A-D scale adapted from intelligence community practices. This helps readers understand the evidentiary weight of our sources.
| Rating | Meaning | Examples |
|---|---|---|
| A | Completely Reliable | Court records, official government documents, legal filings, and other primary legal sources. |
| B | Usually Reliable | Major journalism outlets (NYT, WaPo, AP, Reuters) and official government statements. |
| C | Fairly Reliable | Academic research, peer-reviewed studies, think tank reports with documented methodology. |
| D | Limited Reliability | Primary sources (direct statements, social media) that require additional verification. |
Assessments can become outdated. We review entities on a schedule based on their threat tier, ensuring higher-risk classifications receive more frequent scrutiny.
| Tier | Review Interval | Rationale |
|---|---|---|
T1 | 90 days | Violent extremists require frequent review due to high consequence of error |
T2 | 180 days | Stochastic terrorism cases evolve rapidly and need regular reassessment |
T3 | 1 year | Anti-democratic actors typically show slower behavioral evolution |
T4T5 | 18 months | Lower severity tier with more subjective assessment criteria |
During reviews, we verify that existing evidence remains valid, check for new developments, and assess whether the tier classification is still appropriate. Entities may be upgraded, downgraded, or removed based on review findings.
We acknowledge that all documentation involves judgment. Our commitment is to:
Transparency
We show our sources and reasoning. Every claim links to verifiable evidence.
Consistency
We apply the same standards to all entities regardless of political affiliation.
Correction
We promptly correct errors and maintain a public corrections log.
Humility
Tiers 4 and 5 explicitly acknowledge subjective judgment. We do not claim perfect objectivity.
Questions about our methodology? Contact us.