Four named bodies supply almost all of the wording in circulation: the UN Emergency Relief Coordinator, Human Rights Watch, Amnesty International and Médecins Sans Frontières have each described Israel’s restrictions on humanitarian aid into Gaza as collective punishment, in statements dated between 2009 and 2025. The prohibition they invoke is two sentences long, sits in Article 33 of the Fourth Geneva Convention of 1949, and carries no military-necessity exception. No tribunal has entered a finding of collective punishment on the aid restrictions, and the Rome Statute of 1998 does not list collective punishment among the war crimes it enumerates, while the statutes of two earlier tribunals do.
Key Findings
- Article 33 of the Fourth Geneva Convention (1949) states the rule in two sentences: “No protected person may be punished for an offence he or she has not personally committed. Collective penalties and likewise all measures of intimidation or of terrorism are prohibited.” It admits no derogation for military necessity.
- Four separate provisions, not Article 33, govern relief consignments: Fourth Geneva Convention Articles 23, 55 and 59, and Additional Protocol I (1977) Article 70, which makes relief actions “subject to the agreement of the Parties concerned” while requiring parties to allow and facilitate rapid and unimpeded passage.
- The Rome Statute (1998) does not enumerate collective punishment in Article 8; the Statute of the International Criminal Tribunal for Rwanda (1994) does so at Article 4(b) and the Statute of the Special Court for Sierra Leone (2002) at Article 3(b).
- COGAT (2026) records 1,778,363 metric tons of food entering Gaza between 10 October 2025 and 7 June 2026, its own count, with tonnages self-declared by suppliers rather than weighed at the crossings.
- Becker, Ascone and Troschke (2022, peer-reviewed) coded 1,504 comments on UK media Facebook pages and found 26.9% antisemitic; within those, the apartheid analogy appeared in 5.2% and the Nazi analogy in 4.2%.
Which bodies call the aid restrictions collective punishment?
Four bodies account for nearly all of the phrasing in circulation. The UN Emergency Relief Coordinator, Tom Fletcher, called the halt of humanitarian assistance a “cruel collective punishment” in a 2025 statement issued in that office. Human Rights Watch, Amnesty International and Médecins Sans Frontières apply the same term in their own published statements. None of the four is a court, and none issued a judicial finding.
Each is a different kind of statement, and the difference matters for what a reader can do with it. Fletcher’s is the position of a serving UN humanitarian official, dated 2025. Human Rights Watch and Amnesty International are monitoring organizations with declared advocacy missions; both apply their own reading of the definition to material they gather themselves, and both have said that obstructing relief supplies and depriving civilians of the necessities of survival fall inside the prohibition. Médecins Sans Frontières states the characterization in the title of its own published appeal, “Gaza: Israeli authorities must stop collective punishment.”
The dated specifics are the organizations’ own accounts rather than independently verified tallies. Amnesty International stated in 2025 that aid was “completely cut off” for two months after a ban on aid and commercial goods was reimposed on 2 March 2025; that duration and that date are Amnesty’s own reporting, published in Amnesty’s own statement, with no separate measuring body named for either. Human Rights Watch has anchored its statements to two spans, the period since 7 October 2023 and a preceding closure it describes as 17 years long; those spans, likewise, are the organization’s own framing rather than a tabulated series.
What follows from that is narrow and checkable. Four bodies hold the position, in statements running from 2009 through 2025, and the set of tribunals that have entered the same finding on Gaza aid restrictions is empty. The two categories are not interchangeable: an organization applying a treaty definition to its own gathered material produces a characterization, and a court applying it in adversarial proceedings produces a judgment that binds parties. The remainder of this report reads the instruments those characterizations rest on, the counts the responding party publishes, and the record of what has and has not been decided.
Is it often said the siege is unlawful collective punishment?
It is often said that the closure of Gaza amounts to unlawful collective punishment of its entire population. As a description of who has said it, that is accurate and it is old: Amnesty International’s 2009 legal briefing on the Gaza conflict quoted Article 33 of the Fourth Geneva Convention and stated that the blockade “amounts to collective punishment of its entire population,” 16 years before the current statements.
The wording has been stable across that span. Amnesty International repeated the characterization in statements on the closure through the 2010s and again in 2023 and 2025; Human Rights Watch has used it since October 2023; the UN Emergency Relief Coordinator used it in 2025. The prohibition itself is not novel or marginal in the field. The International Committee of the Red Cross, in its published teaching casebook on the law of armed conflict, treats the ban on collective punishment as a settled rule that applies in international and non-international armed conflict alike.
The base rate is what the repetition does and does not establish. Every item in the sequence is an organizational statement or the position of a serving official, issued outside any adversarial proceeding, applying a definition to material the issuing body assembled itself. Consistency across bodies is a real feature of the record, and it is a feature of the same category of evidence repeated, not of a second category accumulating alongside it. Across the same 2009 to 2026 period, the count of tribunals that have entered a finding of collective punishment in relation to Gaza remains at zero.
That is the shape of the answer to the question as it is usually put. The characterization is widely held, precisely dated, and traceable to named bodies that will state it in their own words. It is also, on the record as it stands, the only category of determination that exists on the point, which is why the rest of this report goes to the instruments themselves.
What does Article 33 of the Fourth Geneva Convention prohibit?
Article 33 of the Fourth Geneva Convention of 1949 states the rule in two sentences: “No protected person may be punished for an offence he or she has not personally committed. Collective penalties and likewise all measures of intimidation or of terrorism are prohibited.” The same article goes on to prohibit pillage and to prohibit reprisals against protected persons and their property. That is the whole of the provision the characterizations rest on.
The rule is repeated in both 1977 Additional Protocols. Additional Protocol I, Article 75(2)(d), lists “collective punishments” among the acts prohibited “at any time and in any place whatsoever, whether committed by civilian or by military agents,” in the article setting out fundamental guarantees. Additional Protocol II, Article 4(2)(b), applies the identical prohibition in armed conflicts not of an international character. The International Committee of the Red Cross records the prohibition as admitting no exception for military necessity and as applying regardless of the conduct of the adversary.
Reach is part of the answer. The Fourth Geneva Convention has 196 states parties on the depositary record kept by the Swiss Federal Council, the widest adherence of any humanitarian-law treaty; Israel ratified it on 6 July 1951. The two Additional Protocols have narrower adherence than the Convention, so a claim resting on Protocol I or II rather than on Article 33 is a claim with a smaller set of bound states behind it, and the distinction is worth keeping when a source cites the Protocols.
Two features of the text govern everything downstream. The first is that the prohibited thing is a penalty: the sentence turns on someone being punished for an offence not personally committed, which is a statement about the character and purpose of a measure, not about how many people a measure affects. The second is that Article 33 says nothing about relief consignments, supply duties, inspection or consent. Those questions are governed by other articles of the same Convention, set out below, and they carry their own tests.
What must be shown before a restriction counts as punishment?
The element that does the work is purpose. Article 33 of the Fourth Geneva Convention (1949) prohibits a penalty imposed on protected persons for an offence they have not personally committed, so the question a decision-maker faces is whether a given measure operates as a sanction on the population rather than as an act with another character. Deprivation that arises incidentally from hostilities is addressed by different rules, on proportionality and precautions, not by Article 33.
That distinction is what separates three claims that are frequently spoken as one. Collective punishment under Article 33 is a penalty imposed for acts not personally committed. Starvation of civilians as a method of warfare is a separate offence with separate elements, enumerated in the Rome Statute of 1998 at Article 8(2)(b)(xxv) as “intentionally using starvation of civilians as a method of warfare by depriving them of objects indispensable to their survival, including wilfully impeding relief supplies as provided for under the Geneva Conventions.” Incidental civilian deprivation resulting from lawful operations is a third thing again, governed by the targeting rules. A finding on any one of them does not carry over to the others.
The forum record follows that split. The International Criminal Court’s Pre-Trial Chamber I issued arrest warrants on 21 November 2024 in the Situation in the State of Palestine, and the charge concerning deprivation of relief in that decision is the Article 8(2)(b)(xxv) starvation offence, because collective punishment as such does not appear in Article 8 of the Rome Statute. A warrant decision is a finding on whether reasonable grounds exist to believe a person committed the charged crimes, not a determination of guilt and not a determination on a charge that was never before the chamber.
The practical consequence for a reader is that the two words in the question, “collective punishment,” name a specific prohibition with a purpose element, and the evidentiary route most available in an international criminal forum runs through a different provision with an intent element about the method of warfare. Two distinct offences, two distinct tests, and sources that use them interchangeably are describing different legal questions.
Which treaty rules govern relief passage into occupied territory?
Four provisions, not Article 33, govern relief. The Fourth Geneva Convention of 1949 sets them out at Articles 23, 55 and 59, and Additional Protocol I of 1977 adds Article 70. Together they specify what an occupying power owes the population, what it must let through, and what conditions it may attach, which is the terrain on which any aid dispute is actually decided.
Article 23 obliges each party to allow free passage of consignments of medical supplies, and of essential foodstuffs, clothing and tonics intended for children under fifteen, expectant mothers and maternity cases. The same article attaches conditions in its own text: the permitting party may satisfy itself that consignments are not diverted from their destination, that control is effective, and that no definite advantage accrues to the military efforts or economy of the enemy.
Article 55 places the supply duty on the occupier directly. It provides that the occupying power has the duty of ensuring the food and medical supplies of the population “to the fullest extent of the means available to it,” and that it “should, in particular, bring in the necessary foodstuffs, medical stores and other articles if the resources of the occupied territory are inadequate.” Article 59 then provides that where the population of an occupied territory is inadequately supplied, the occupying power “shall agree to relief schemes on behalf of the said population, and shall facilitate them by all means at its disposal,” while permitting it to search consignments and to prescribe technical arrangements for their passage.
Additional Protocol I, Article 70, carries the consent question that most disputes turn on. Relief actions for a civilian population “shall be undertaken, subject to the agreement of the Parties concerned in such relief actions,” and the parties “shall allow and facilitate rapid and unimpeded passage of all relief consignments, equipment and personnel.” Four provisions, then, and each supplies a different test: free passage with conditions, a duty to ensure supply, a duty to agree and facilitate, and a consent-plus-facilitation rule. A source that argues the whole question through Article 33 alone has left all four out.
Does Israel’s record address the duty to ensure food supplies?
Israel’s own bodies publish tonnage counts and no counts at all for the precautions they describe. COGAT, the Israeli Ministry of Defense unit that coordinates civil activity in the territories, records 1,778,363 metric tons of food entering Gaza between 10 October 2025 and 7 June 2026, its own count, which it sets at nearly threefold the World Food Programme’s monthly requirement benchmark. Every figure in this section is a party’s own reporting, none of it independently verified.
The other counts come from the same category. The Israel Defense Forces states that over 1.9 million tons of aid were delivered between the start of the war and 19 August 2025, including 7,800 tons by the aerial route, and that it oversaw 23,000 tons in the week ending on that date. The Israeli Ministry of Foreign Affairs reported over 70,000 tons of aid entering by 16 December 2023, including over 6,500 tons of medical supplies. Israel’s Permanent Mission to the Multilateral Organizations in France reported more than 106,000 aid trucks carrying over two million tons as of 14 September 2025, and more than 12,000 trucks since May 2025 with 80% of them carrying food. COGAT notes that its tonnages rest on self-declaration by humanitarian organizations and private providers rather than weighing at the crossings.
On the precautions themselves the record is different in kind. The Israel Defense Forces states that it warns civilians by telephone, text message and leaflet, uses a non-lethal munition before striking a building, aborts strikes when civilians are present, and published a numbered evacuation-area map, and it states that Hamas sites military infrastructure in residential buildings, hospitals, mosques and schools. It publishes no count of warnings issued, roof-knocks performed or strikes aborted, so no magnitude attaches to any of it. The volume dispute that these tonnage counts sit inside is treated separately in the Institute’s report on the starvation policy claim.
Has any tribunal adjudicated the collective punishment charge?
Zero tribunals have entered a finding of collective punishment on the Gaza aid restrictions.
The International Court of Justice delivered an advisory opinion on Israeli policies and practices in the occupied Palestinian territory on 19 July 2024, and issued provisional measures orders in South Africa v. Israel on 26 January, 28 March and 24 May 2024, the later two addressing humanitarian access; none of the four is a finding on collective punishment, and an advisory opinion is by its nature non-binding.
The instrument record explains part of the gap. Collective punishment is prohibited by treaty but is not enumerated as a crime within the jurisdiction of the International Criminal Court, whose Rome Statute of 1998 lists war crimes exhaustively in Article 8 without it. Two earlier tribunal statutes did enumerate it, which is why the charge has a litigation history elsewhere and none here.
| Instrument | Collective punishment provision | Source type |
|---|---|---|
| Fourth Geneva Convention, 1949 | Article 33, prohibited | Treaty text (official record) |
| Additional Protocol I, 1977 | Article 75(2)(d), prohibited | Treaty text (official record) |
| Additional Protocol II, 1977 | Article 4(2)(b), prohibited | Treaty text (official record) |
| ICTR Statute, 1994 | Article 4(b), enumerated as a crime | Statute text (official record) |
| Special Court for Sierra Leone Statute, 2002 | Article 3(b), enumerated as a crime | Statute text (official record) |
| Rome Statute, 1998 | Not enumerated in Article 8 | Treaty text (official record) |
The consequence is structural rather than evidentiary. A prohibition can be firmly established in the law of armed conflict and still have no forum in which it is charged as such, and that is the position of Article 33 before the International Criminal Court. Where deprivation of relief is charged there, it is charged under Article 8(2)(b)(xxv) as starvation of civilians as a method of warfare, with the intent element that provision carries. Israel is not a party to the Rome Statute, a fact of the ratification record that bears on jurisdiction and is contested between the parties to the proceedings.
Where do analysts dispute the collective punishment label?
Three published Israeli and pro-Israel research bodies contest either the characterization or the methods behind it, and the clearest quantitative example is the Jerusalem Center for Security and Foreign Affairs, which in 2025 derived a ceiling of at most 41,000 civilian war-related deaths in Gaza by subtracting estimated natural deaths and Israel’s count of fighters killed from the Gaza Ministry of Health total. The Center states on the same page that neither input has been independently confirmed.
That figure is the publisher’s own arithmetic on inputs supplied by parties to the conflict, and it is presented as a ceiling rather than a count. It is cited here for what it shows about the dispute rather than as a settled total: the contested step is not the treaty text but the attribution of purpose, and a lower civilian figure is offered by its author as evidence against reading the pattern of restriction as a penalty imposed on a population.
The Begin-Sadat Center for Strategic Studies publishes running war totals from the other direction, reporting 1,664 Israelis confirmed killed and 17,809 wounded as of 5 September 2024, and 1,184 killed on 7 October 2023, with no official body credited for the first two. NGO Monitor, a monitoring organization with a declared advocacy mission, directs its 2024 annual report at the evidentiary methods of the organizations issuing determinations, reporting that it identified 157 groups in the United States and 111 in Canada within what it describes as an anti-Israel network, with over 3,000 and over 1,800 connections respectively; the report does not state the inclusion criteria or the mapping method.
Three points are in genuine dispute across these publications and the organizational statements in the first section: whether a pattern of restriction evidences a punitive purpose or a set of inspection, diversion and consent judgments under Articles 23 and 59; whether diversion of consignments has been measured or assumed by either side; and whether an organizational statement carries the evidentiary weight of a determination. The first of those is the element Article 33 actually turns on, and no source on either side of it publishes a measured finding on purpose.
How does the punishment framing appear in measured discourse?
Israel-related framings appear inside measured antisemitic comment corpora at rates that have been counted. Becker, Ascone and Troschke (2022, peer-reviewed) found that 26.9% of comments on the Facebook pages of leading UK media outlets reacting to the May 2021 escalation were antisemitic against an IHRA-based scheme.
Becker, Ascone and Troschke coded 1,504 UK comments in that sample, and under the same IHRA-based scheme they also found 12.6% of 1,500 French comments and 13.6% of 1,520 German comments to be antisemitic.
Within the UK antisemitic comments, the study reports the shares carrying each framing. The categories are not mutually exclusive and the corpus is a keyword-bounded sample of comments on specific pages during one escalation, not a measurement of public discourse generally.
Survey work measures associations, not causes, and the authors state that limit themselves. Enstad (2026, peer-reviewed, abstract-sourced) ran two nationally representative Norwegian surveys, Study 1 with 1,575 respondents and a preregistered replication with 1,653, and found that endorsement of Holocaust inversion, the belief that Israel treats Palestinians as badly as Jews were treated during the Second World War, was associated with justification of harassment and violence against Jews and with refusal to take a stance against it. Ozer, Obaidi and Bergh (2025, peer-reviewed, abstract-sourced) ran two British experiments, with 354 and 490 participants, priming war-in-Gaza material and identifying two separate pathways to anti-Jewish hostility, one through social dominance orientation and one through identity fusion with the Palestinian people.
Harber, Bulska, Malloy and Vila (2026, peer-reviewed, abstract-sourced) report across three preregistered studies that conspiracy beliefs implicating Israel and Zionists mediated the association between antisemitism and anti-Israel attitudes, with the model accounting for over 55% of the variance in anti-Israel attitudes, while conspiracy beliefs about Jews unrelated to Israel and a general conspiratorial mindset did not. These are correlational designs measuring attitudes held by survey respondents. None measures the legal characterization discussed in this report, and none measures the people or organizations issuing it. Related measurement is collected in the Institute’s report on anti-Zionism and antisemitism.
Methodology and limitations
This report measures three things: which named bodies have applied the term collective punishment to Israeli restrictions on aid into Gaza and when, what the governing treaty and statute texts say, and what counts each party to the events publishes about aid entering the territory. It does not measure conditions inside Gaza, and no figure here is an Institute measurement of aid volume, deprivation or casualty.
Sources are used by type. Treaty and statute texts are quoted from the instruments themselves and are the highest tier used here: the Fourth Geneva Convention of 1949, the two Additional Protocols of 1977, the Rome Statute of 1998, and the statutes of the International Criminal Tribunal for Rwanda and the Special Court for Sierra Leone. Statements by Human Rights Watch, Amnesty International, Médecins Sans Frontières and the UN Emergency Relief Coordinator are reported as those bodies’ own characterizations, named and dated, and their internal figures, including Amnesty’s two-month duration and the 2 March 2025 date, are attributed to them rather than carried as measured quantities. Israeli government and military figures, from COGAT, the Israel Defense Forces, the Military Advocate General’s Corps and the Ministry of Foreign Affairs, are the reporting party’s own counts of its own activity, published by a party to the events, and are not independently verified; COGAT states that its tonnages rest on supplier self-declaration rather than weighing at the crossings, and the Ministry and Defense Forces figures are written as floors, so “over 1.9 million tons” cannot be restated as an exact total.
Windows and scopes differ across the Israeli counts, which is why the chart in the sixth section carries all three in its labels: one covers all aid to December 2023, one all aid to August 2025, and one food alone from October 2025. They are not a series and must not be read as one.
Three studies cited in the discourse section (Enstad, 2026; Ozer, Obaidi and Bergh, 2025; Harber, Bulska, Malloy and Vila, 2026) are read from published abstracts, so the reported detail is limited to what the abstract states. All four studies in that section report associations, not causal effects, and the Becker corpus is a keyword-bounded sample of comments on specific media pages during a single escalation. Where a determination on collective punishment is absent from the tribunal record, that absence is reported as a fact about the record, not as an assessment of any claim.
Conclusion
The question turns on a phrase from 1949: whether people are being punished for an offence they have not personally committed. On the record as it stands, four named bodies say that is what the aid restrictions amount to, in statements running from Amnesty International’s 2009 briefing to the UN Emergency Relief Coordinator’s 2025 wording, and no tribunal has entered that finding.
The rest of the record is sharper than the debate around it usually allows. Article 33 is two sentences and admits no military-necessity exception, but it says nothing about consignments, and the rules that do, Articles 23, 55 and 59 and Additional Protocol I Article 70, each carry their own test of duty, inspection and consent. The element Article 33 actually turns on is purpose, and no source on either side publishes a measured finding on purpose. The forum most associated with the charge does not list it: the Rome Statute omits collective punishment where the Rwanda and Sierra Leone statutes enumerated it. The responding party publishes tonnage in the millions and publishes no count whatever of the warnings, roof-knocks or aborted strikes it describes. The organizations issuing determinations and the institutes contesting them disagree about attribution of purpose, about diversion, and about what an organizational statement is worth as evidence.
Meanwhile the vocabulary travels outside the legal argument entirely. Becker, Ascone and Troschke counted the Israel-related framings inside antisemitic comments on mainstream British media pages, where the sole-guilt frame appeared in 27.9% and the apartheid analogy in 5.2%, and Enstad’s two Norwegian samples found endorsement of Holocaust inversion associated with justification of aggression against Jews. A term of art with a purpose element and a treaty article behind it becomes, in general speech, a word for deliberate cruelty toward a people. When a legal characterization crosses that line before any tribunal has applied it, what does the word carry with it, and is that migration one of the inputs that feeds what those studies were counting?
Frequently Asked Questions
Does the prohibition on collective punishment apply in a conflict that is not between states?
Yes. Additional Protocol II of 1977, Article 4(2)(b), prohibits collective punishments in armed conflicts not of an international character, and Additional Protocol I, Article 75(2)(d), prohibits them “at any time and in any place whatsoever.” Adherence to the Protocols is narrower than to the Fourth Geneva Convention itself, so the set of bound states differs.
Who counts as a protected person under the Fourth Geneva Convention?
Article 4 of the Fourth Geneva Convention (1949) defines protected persons as those who, at a given moment and in any manner whatsoever, find themselves in the hands of a party to the conflict or occupying power of which they are not nationals. Article 33 runs only to that category, which is why the status question precedes the punishment question in any argument built on the article.
Can an occupying power inspect or refuse a relief consignment?
Article 23 of the Fourth Geneva Convention lets a permitting party satisfy itself that consignments are not diverted, that control is effective, and that no definite advantage accrues to the adversary’s military effort or economy, and Article 59 permits search of consignments and technical arrangements for their passage. Article 59 pairs those permissions with an obligation to agree to relief schemes and facilitate them by all means at its disposal.
Why is starvation, rather than collective punishment, the charge available at the International Criminal Court?
Because Article 8 of the Rome Statute (1998) enumerates war crimes exhaustively and does not include collective punishment, while Article 8(2)(b)(xxv) does enumerate intentionally using starvation of civilians as a method of warfare, including wilfully impeding relief supplies. The two offences have different elements, so a determination on one is not a determination on the other.
Has Israel’s military opened investigations into conduct during the Gaza war?
The Military Advocate General’s Corps reported 74 criminal investigations opened as of 3 August 2024, including 44 concerning the death of detainees and 3 concerning destruction of civilian property without military necessity, alongside over 1,000 incidents referred for factual assessment. These are the Corps’ own counts of investigations opened, not findings or convictions, and the Corps states that civilian death or injury in armed conflict “in most cases do not by themselves indicate a reasonable suspicion of criminal conduct.”
Does Israel publish figures for the warnings it says it issues?
No. The Israel Defense Forces describes telephone calls, text messages, leaflets, roof-knocking and a numbered evacuation-area map, and it publishes no count of warnings issued, roof-knocks performed or strikes aborted. Any magnitude attached to those measures would come from outside the publishing body’s own record.
How often does Israel-related framing require surrounding context to read as antisemitic?
Becker, Ascone and Troschke (2022) reported that thread context was needed to infer the antisemitic meaning in 38.7% of the UK antisemitic comments they coded, about 62% of the French and 48.3% of the German. That share is one reason keyword-based counts and context-coded counts of the same material do not agree.
Sources
- Becker, Ascone and Troschke, 2022. Antisemitic Comments on Facebook Pages of Leading British, French, and German Media Outlets. Humanities and Social Sciences Communications. DOI 10.1057/s41599-022-01337-8. Peer-reviewed.
- Begin-Sadat Center for Strategic Studies, 2025. The Gaza Terror Offensive, 1 August to 5 September 2024. besacenter.org. Research institute, not peer-reviewed.
- COGAT (Coordinator of Government Activities in the Territories), 2026. Humanitarian Situation Report During the Ceasefire. gaza-aid-data.gov.il. Official record, Israeli Ministry of Defense unit, self-reported.
- Diplomatic Conference of Geneva, 1949. Geneva Convention Relative to the Protection of Civilian Persons in Time of War (Fourth Geneva Convention), 12 August 1949, Articles 4, 23, 33, 55 and 59. United Nations Treaty Series, vol. 75, No. 973; ratifications recorded by the Swiss Federal Council as depositary. Official record, treaty text.
- Diplomatic Conference on the Reaffirmation and Development of International Humanitarian Law, 1977. Protocol Additional to the Geneva Conventions of 12 August 1949, and Relating to the Protection of Victims of International Armed Conflicts (Protocol I), Articles 70 and 75. UNTS vol. 1125, No. 17512. Official record, treaty text.
- Diplomatic Conference on the Reaffirmation and Development of International Humanitarian Law, 1977. Protocol Additional to the Geneva Conventions of 12 August 1949, and Relating to the Protection of Victims of Non-International Armed Conflicts (Protocol II), Article 4. UNTS vol. 1125, No. 17513. Official record, treaty text.
- Enstad, 2026. Holocaust Inversion and Justification of Anti-Jewish Aggression: Evidence from Two National Surveys in Norway. American Psychologist. DOI 10.1037/amp0001704. Peer-reviewed, abstract-sourced.
- Harber, Bulska, Malloy and Vila, 2026. Antisemitism, Conspiracy Beliefs, and Anti-Israel Attitudes. American Psychologist. DOI 10.1037/amp0001635. Peer-reviewed, abstract-sourced.
- Israel Defense Forces, 2025. Clearing Up the Misconceptions: Everything You Need to Know About the Humanitarian Aid Situation in Gaza. idf.il. Official record, self-reported by a party to the conflict.
- Israel Defense Forces. Stated Precautions in Gaza and the Human-Shields Position. idf.il. Official record, a party’s own stated position; publishes no counts.
- Israel Defense Forces, Military Advocate General’s Corps, 2024. Addressing Alleged Misconduct in the Context of the War in Gaza, published 24 February 2024, updated 3 August 2024. idf.il. Official record, self-reported.
- Israel Ministry of Foreign Affairs, 2023. Israel-Hamas Conflict 2023: Humanitarian Efforts. gov.il. Official record, self-reported.
- Israel Ministry of Foreign Affairs, Permanent Mission of Israel to the Multilateral Organizations in France, 2025. Israel’s Objectives and Current Operation, updated 17 September 2025. embassies.gov.il. Official record, self-reported.
- Jerusalem Center for Security and Foreign Affairs, 2025. The Death of Genocide. jcfa.org. Research institute, not peer-reviewed.
- NGO Monitor, 2025. Annual Report 2024. ngo-monitor.org. Monitoring organization, advocacy mission, not peer-reviewed.
- Ozer, Obaidi and Bergh, 2025. The Impact of Globalized Conflicts: Examining Attitudes Toward Jews Among Britons in the Political Context of the War in Gaza. International Journal of Intercultural Relations, vol. 107. DOI 10.1016/j.ijintrel.2025.102184. Peer-reviewed, abstract-sourced.
- United Nations, 1998. Rome Statute of the International Criminal Court, adopted 17 July 1998, Article 8. UNTS vol. 2187, No. 38544. Official record, treaty text.
- United Nations and Government of Sierra Leone, 2002. Statute of the Special Court for Sierra Leone, Article 3, annexed to the Agreement of 16 January 2002. Official record, statute text.
- United Nations Security Council, 1994. Statute of the International Criminal Tribunal for Rwanda, Article 4, annex to Resolution 955 (1994). Official record, statute text.
