Methodology

Purpose

This is a public archive of alleged war crimes, crimes against humanity, and serious violations of international law from the second Trump administration (January 20, 2025 onward). It is a project of the Lilac Party.

It is not a news feed. Every entry has an incident date, an update date, a legal posture label, and a source trail you can check yourself.

Legal Disclaimer

This archive documents allegations, not adjudicated findings. The title is editorial framing — it is not a blanket legal conclusion that every record here constitutes an adjudicated war crime.

No person named in this archive has been convicted of any war crime by any tribunal. When we use terms like "war crime" or "crime against humanity," we are drawing on analysis from recognized authorities — Human Rights Watch, Amnesty International, Lawfare, Just Security, and others — and citing specific provisions of international law. That is credible legal analysis, not a verdict.

Everything here comes from public reporting, court filings, government records, and published legal analyses. Every incident page has source citations you can check.

Each incident page separates:

reported facts from cited public sources
official acts acknowledged by the government
claims made in litigation or public advocacy
judicial findings where a court order or ruling is cited
the publication's own severity assessment

If a page does not cite a ruling or operative official record for a legal conclusion, readers should treat legal terms on the site as documented analysis or allegation, not as a final adjudication.

Incident Selection Criteria

What Gets In

Every incident has to clear these bars:

Publicly sourceable: backed by reporting, public filings, government records, or other verifiable material
Dateable: tied to a specific date or clearly documented time period
Classifiable: fits the archive's category, severity, and legal-posture system
Explainable: can be described plainly without overstating what the record supports
Auditable: comes with a source trail you can inspect

Qualifying Incidents

An incident qualifies for the archive if the conduct described, taken at face value from credible sources, maps to one or more of the following:

Violations of the Geneva Conventions or their Additional Protocols
Crimes against humanity as defined in the Rome Statute (Article 7)
War crimes as defined in the Rome Statute (Article 8)
Acts of genocide as defined in the Genocide Convention and Rome Statute (Article 6)
Violations of the Convention Against Torture
Violations of the principle of non-refoulement under international refugee law
Serious violations of constitutional due process protections where the conduct also implicates international law obligations
Material facilitation of any of the above (Rome Statute Article 25(3)(c))

Domestic policy disputes, ordinary political misconduct, and actions that are objectionable but do not implicate international humanitarian or human rights law are outside scope, no matter how serious they may be in other contexts.

What Does Not Get In

Domestic-only legal disputes that do not touch international law obligations
Unsubstantiated claims without public sourcing, regardless of how credible the claimant
Anonymous tips or information from protected sources that cannot be independently verified through public material
Speculation about future conduct — the archive documents what has happened, not what might happen
Incidents from prior administrations — the scope is January 20, 2025 onward

Borderline Cases

Some incidents sit at the edge of what qualifies. Our approach:

When credible sources disagree on whether conduct rises to the level of an international law violation, we include the incident and note the disagreement
When an incident involves domestic law violations that also implicate international obligations (e.g., deportation orders that may violate non-refoulement), we include it with clear framing of both dimensions
When the legal theory is novel or untested, we include the incident if the factual record is strong and at least one recognized legal authority has articulated the theory in a published analysis
Borderline incidents are tagged with lower severity ratings and more cautious legal posture labels until the record develops

Source Hierarchy

Not all sources carry equal weight. The archive uses a tiered system when evaluating and presenting evidence.

Tier 1 — Primary Documents

The strongest evidence. These are the records themselves:

Court orders, judicial opinions, and rulings
Executive orders, presidential proclamations, and official memoranda
Congressional testimony and official hearing transcripts
Government agency records, reports, and press releases
International tribunal filings and decisions
Treaty text and formal legal instruments

When primary documents exist for an incident, they are cited directly and listed separately on the incident page.

Tier 2 — Investigative Journalism

Original reporting from outlets with established editorial standards, fact-checking processes, and track records of accuracy:

Major wire services: Reuters, Associated Press
Papers of record: The New York Times, The Washington Post, The Guardian
Investigative units with demonstrated methodology: ProPublica, The Intercept, BBC Investigations

Tier 2 sources carry high weight when they include direct evidence (documents obtained, on-the-record interviews with participants, field reporting from the scene).

Tier 3 — Analytical Legal Commentary

Published legal analysis from recognized authorities:

International human rights organizations: Human Rights Watch, Amnesty International, International Committee of the Red Cross
Legal analysis platforms: Lawfare, Just Security, Opinio Juris
Academic legal scholars publishing in their area of expertise
Bar association reports and formal legal opinions

Tier 3 sources are used primarily for legal classification — identifying which provisions of international law may apply to documented facts. They are not treated as independent factual sources unless they include original investigation.

Tier 4 — Single-Source Reporting

Reporting from a single outlet without corroboration, or from outlets with less established editorial processes. This includes:

Local news coverage of a specific event
Single-reporter stories without documentary evidence
Social media posts from officials or witnesses (used only as supporting evidence, never as sole basis)

Tier 4 sources can support an incident entry, but an incident built solely on Tier 4 material will carry an "Unverified" verification status until corroborated.

Minimum Source Requirements

Every incident requires at least one source. There is no incident page without a citation. The verification status (see below) reflects the depth and diversity of the source base, not just the count.

Verification Methodology

Every incident carries a verification status reflecting how thoroughly the factual claims have been checked against independent material.

Independently Verified

The highest tier. An incident reaches this status when:

Two or more unrelated sources have corroborated the core facts
Those sources used distinct methodologies — for example, field reporting combined with satellite imagery, or government records combined with an independent NGO investigation
The sources do not share an upstream origin (i.e., they are not both reporting on the same single leak or press release)

"Independently verified" means the factual foundation has been confirmed through multiple independent lines of evidence. It does not mean a court has ruled on legality.

Cross-Referenced

The middle tier. An incident reaches this status when:

Two or more published sources corroborate the core facts
The sources may share an upstream origin (e.g., both reporting on the same government filing, both citing the same NGO report)
The corroboration is factual, not merely editorial agreement

Cross-referenced incidents have a solid factual base but may lack the methodological diversity of independently verified entries.

Unverified

The baseline. An incident carries this status when:

Only a single source has been identified
Multiple sources exist but have not yet been cross-checked
The incident was recently added and verification is in progress

"Unverified" does not mean unreliable. It means the verification process is incomplete. Many incidents enter the archive at this tier and are upgraded as additional sources are identified.

Classification Methodology

Severity

Severity is our editorial assessment of how grave the rights and rule-of-law risk is. It is not a criminal charge or tribunal judgment.

Critical Rights and Rule-of-Law Concern: unusually grave questions involving unlawful detention, refoulement, executive defiance, or severe due-process breakdowns
Serious Rights Violation: serious constitutional, statutory, or international-law concerns supported by strong public evidence
Major Abuse of Power: substantial institutional abuse, dismantlement, or coercive misuse of public authority
Significant Democratic Concern: serious democratic-accountability or rule-of-law harms that may not yet rise to the levels above

Severity ratings are assigned based on the factual record and the legal analysis available at the time of publication. They may be revised upward or downward as the record develops.

Legal Posture

Every incident page carries a posture label reflecting its current legal and evidentiary status:

Reported record: the page relies on reporting or public records, without citing a final judicial finding on the core legal question
Active litigation: courts are actively considering legality, remedy, or compliance issues tied to the incident
Judicial finding: the page cites a court ruling or operative order directly bearing on the contested conduct
Official executive action: the act itself is publicly acknowledged by the government, even if the legal critique remains contested

War Crime Classification

Separate from severity and legal posture, each incident carries a classification reflecting how close the documented conduct is to a formal finding of illegality under international law:

Confirmed: a formal legal finding by an international tribunal, the International Criminal Court, or an equivalent judicial body has determined that the conduct constitutes a violation of international humanitarian or human rights law. This is the highest bar and requires an actual adjudication, not merely an investigation or indictment.
Probable: multiple independent credible sources corroborate factual elements that map directly to specific articles of the Rome Statute, Geneva Conventions, or other binding instruments. Recognized legal authorities (Tier 3 sources) have published analyses reaching the same conclusion. The evidentiary record would support a formal proceeding, but no adjudication has occurred.
Potential: credible reporting indicates conduct that could constitute a violation of international law if the reported facts are accurate. The legal theory is recognized, but the factual record may be incomplete, contested, or based on fewer independent sources. This classification signals that the conduct warrants serious scrutiny, not that a violation has been established.
Enabling: the documented conduct materially facilitates violations committed by others, within the meaning of Rome Statute Article 25(3)(c) — aiding, abetting, or otherwise assisting in the commission or attempted commission of a crime. This includes providing the means, removing legal safeguards, or creating the conditions that make violations possible. The facilitation must be documented, not speculative.

Classification may change over time as the factual record develops, legal proceedings advance, or new analysis is published.

Editorial Independence

Who Maintains This

This archive is maintained by the Lilac Party. There is no corporate parent, no advertising revenue, no grant funding from organizations with a stake in the subject matter, and no editorial board drawn from government or political party affiliations.

Editorial Process

Incidents are researched, drafted, and reviewed before publication. The process:

Identification: an event or pattern of conduct is identified through monitoring of public sources
Source gathering: primary documents and reporting are collected and organized by tier
Drafting: the incident page is written, separating facts from analysis and clearly attributing claims to their sources
Classification: severity, legal posture, verification status, and war crime classification are assigned based on the source base
Review: the draft is reviewed for accuracy, tone, and compliance with this methodology before publication
Publication: the incident goes live with its full source trail

Conflicts of Interest

The Lilac Party is a political organization. That is a relevant disclosure. This archive exists because the Lilac Party believes the documented conduct warrants public scrutiny and a permanent record.

We do not claim to be neutral. We claim to be accurate. Every factual assertion is sourced. Every legal characterization is attributed. Every severity rating is an editorial judgment clearly labeled as such. Readers can check every citation and reach their own conclusions.

There are no financial conflicts. No funder, advertiser, or donor influences what gets published, how it is classified, or when it is updated.

Corrections Policy

Errors

When a factual error is identified — a wrong date, a misattributed quote, an incorrect legal citation — it is corrected as soon as the error is confirmed. Corrections are logged in the incident's update history with a clear note of what changed and why.

We do not silently fix errors. If the correction is material (changes the meaning, alters the factual basis, or affects the classification), it gets an update log entry visible on the incident page and in the Updates feed.

What Triggers a Correction vs. an Update

Correction: the original text was wrong. A fact was misstated, a source was misattributed, a date was incorrect. The update log entry is labeled as a correction.
Update: new information has emerged. A court ruled. New reporting was published. The classification changed based on new evidence. The update log entry describes what changed and what prompted the change.
Minor edit: typo fixes, formatting changes, link repairs, wording improvements that do not alter meaning. These may not get individual update log entries.

Timeliness

Corrections to factual errors are prioritized and published within 24 hours of confirmation. Updates based on new developments are published as soon as the new material has been reviewed and the page has been revised to reflect it accurately.

Limitations and Biases

This archive has real limitations. Being honest about them is part of the methodology.

Scope

US-focused: this archive covers one country's government. Other governments commit violations of international law. This archive does not cover them. That is a scope decision, not a statement about relative severity.
One administration: the archive begins January 20, 2025. Prior administrations committed acts that could be documented under the same framework. This archive does not cover them. The temporal scope is a practical choice — documenting everything is beyond our capacity.
English-language sources: the source base skews heavily toward English-language reporting and legal analysis. This means incidents covered primarily in non-English media may be underrepresented or delayed.

Severity ratings are editorial judgments. Two reasonable people applying the same framework to the same facts could reach different severity assessments. We are transparent about the criteria, but the ratings are not mechanical outputs — they involve judgment.
Legal classification depends on available analysis. If no recognized legal authority has published analysis mapping documented conduct to specific legal provisions, we may underclassify an incident or delay its publication.
Verification takes time. The archive is a living document. At any given moment, some incidents may carry lower verification status than the underlying facts warrant, simply because the cross-referencing process is ongoing.
Selection bias is real. Incidents that receive heavy media coverage are more likely to be identified and documented quickly. Incidents affecting populations with less media access may be underrepresented.

What This Archive Cannot Do

It cannot compel legal action
It cannot substitute for a formal investigation with subpoena power
It cannot access classified material, sealed court records, or non-public government documents
It cannot protect sources — we publish only what is already public
It cannot provide legal advice to any person or organization

How to Use This Archive

For Researchers

The archive is designed to be citable. Every incident has a stable URL, a structured data export, and a source trail. The data exports section describes the machine-readable formats available.

When citing this archive in academic work, cite the specific incident page (not the homepage) and note the verification status and classification at the time of access. These may change as the record develops.

The archive is a secondary source. For primary evidence, follow the source citations on each incident page to the underlying documents and reporting.

For Lawyers

Each incident page separates factual claims from legal analysis and identifies the source of each. Legal posture labels indicate whether courts have weighed in. War crime classifications reference specific provisions of international law.

This archive is not a legal brief and should not be cited as one. It is a research tool that organizes public material. The underlying sources — court filings, legal analyses, government records — are what belong in legal submissions.

For Journalists

The archive can serve as a starting point for reporting, not a substitute for it. The source trail on each page points to the underlying material. The structured data exports can help identify patterns across incidents.

We distinguish between what we have independently verified, what we have cross-referenced, and what remains unverified. If you are reporting on an incident that carries "unverified" status here, do your own verification before publishing.

For the Public

Read the incident pages. Check the sources. The archive is written in plain language precisely so that you do not need a law degree to understand what is documented here.

The severity ratings and classifications are our assessment. You may disagree. The source citations are there so you can form your own view based on the same evidence.

If something is missing, it may not mean it did not happen — see What Is Missing below.

Data Exports

The archive publishes machine-readable data so you can audit, cite, or build on it without scraping HTML.

/archive.json — full structured archive
/archive.csv — tabular export of the same data
incident/[slug].json — structured record for a single incident
/updates.json — the update ledger

These exports contain only what is already visible on the public site. No private notes, unpublished drafts, or source-protection data.

What This Site Is Not

This site is not:

a court
a criminal indictment
legal advice
a substitute for a full investigative file
a secure channel for sensitive source contact

What Is Missing

This archive is incomplete. A missing incident does not mean it did not happen or did not matter. It may mean:

the public sources are not yet strong enough to meet our bar
the incident has not been converted from our research notes into a publishable record
we do not yet have a good category or collection structure for it
publishing it would require source-sensitive material we cannot make public

If you have information, see Source Safety. We do not currently offer SecureDrop or an equivalent anonymous submission system.