What is UNRWA — and is it infiltrated by Hamas?¶
Short answer: UNRWA is the UN agency set up in 1949 for Palestinian refugees. It's unusual in two ways: it serves only one refugee group, and it passes refugee status down to their descendants. On the "infiltrated by Hamas" claim, keep three separate things straight:
- Some individual staff took part in terrorism. This is documented — and the UN itself has admitted part of it.
- A US inspector general has flagged 108 current or former staff for taking part in October 7 or belonging to Hamas's armed wing. That's a serious, three-figure law-enforcement number; one case is finished and upheld, and the rest await a decision.
- No court has ruled on any of it. The two most recent official records even point opposite ways: the World Court (ICJ, October 2025) said the broad claims weren't backed by the evidence it had, while the US inspector general (June 2026) made its referrals.
Every finished review has found real cases; none has cleared the whole agency. And UNRWA's deepest problems are built into how it was designed — they'd still be there even with a perfectly clean staff.
What is documented [each item tagged]¶
- October 7 participation by staff: Israel accused 19 of UNRWA's ~13,000 Gaza employees of taking part in the attack. The UN's own Office of Internal Oversight Services investigated. It terminated nine staff whose cases showed evidence that they "may have been involved." Nine other cases closed for lack of evidence [VERIFIED via the UN OIOS investigation outcome, August 2024, announced by UNRWA itself; caveat noted both ways — OIOS could not independently authenticate the underlying Israeli intelligence, which Israel largely withheld]
- Hostages held by UNRWA employees: released hostages said they were held in the homes of UNRWA staff, including a teacher. They included an 84-year-old woman and a hostage held ~50 days [DOCUMENTED via released-hostage testimony, contemporaneous press]
- Hamas facilities under and inside UNRWA premises: the IDF found a tunnel and Hamas data center directly beneath UNRWA's Gaza City headquarters. It drew power from the compound's own electrical supply (shown to journalists, February 2024; UNRWA said it had no knowledge and had vacated the compound) [DOCUMENTED via press inspection of the site; UNRWA response on record]
- Weapons in UNRWA schools in earlier wars: in 2014, UNRWA itself found rockets stored in its schools three times, and it condemned this in public. A UN Board of Inquiry confirmed armed-group use of premises [VERIFIED via UNRWA's own 2014 statements and the UN Board of Inquiry report]
- The cost to UNRWA's own staff: more than 310 UNRWA staff killed in the Gaza war — the highest UN personnel death toll in any conflict in the organization's history [DOCUMENTED via UNRWA/UN reporting]. At the same time, the agency remained the largest humanitarian operator in Gaza, including its role in the WHO-coordinated polio campaign [DOCUMENTED via WHO/UN records]
Aid diversion: the separate question, both columns [cross-reference: Document 38]¶
The staff/labor question above is a separate matter from the financial-diversion question — whether Hamas captures aid goods. People often mix the two up. Both sides of the diversion record are documented:
- Real diversion, documented: Hamas taxed commercial goods, seized shipments, and controlled distribution. On November 16, 2024 the UN reported 98 of 109 aid trucks lost or hijacked inside Gaza in a single day [VERIFIED via UN reporting]. Israeli/COGAT figures alleged ~25% diversion early in the war [SINGLE-SOURCE — Israeli-sourced]
- Inflation of the claims, equally documented: the viral claim "the UN says 87% of Gaza aid is looted by Hamas" is a distortion. The UN blamed most convoy losses on hungry civilians offloading trucks, not on Hamas capture. And a July 2025 USAID report looked at 156 incidents and found no evidence of widespread Hamas theft of US-funded aid [VERIFIED correction]
- The distinction that keeps it honest: direct diversion means Hamas physically taking goods. It is documented in specific incidents, though the scale is genuinely disputed. That is a different kind of evidence from fungibility — the idea that aid frees up Hamas's own resources. Fungibility is a structural argument. It is logically valid, but it is not seized-truck evidence. Honest reporting does not present the second as if it were the first [ANALYTICAL FRAMEWORK]
The 2026 escalation: the US inspector general's referrals [each element tagged]¶
- June 5, 2026: USAID's Office of Inspector General is a US law-enforcement body that is independent by statute. It referred 101 current or former UNRWA staff (108 cumulative) to the State Department for suspension/debarment. The stated grounds: "participation in the October 7, 2023, terrorist attacks and/or affiliation with the military wing (al-Qassam Brigades) of Hamas." The findings named specific roles: deputy school principals identified as al-Qassam company and platoon commanders; a teacher in Hamas military intelligence tracking explosive-device assignments; a teacher ordered to deliver two anti-tank missiles on October 7; and a school principal in a Hamas weapons-manufacturing unit whose school sat over a tunnel shaft with three anti-tank positions [VERIFIED — the OIG's own investigative summary]
- What the referral standard means — the ladder printed in full: an OIG referral is an investigator's recommendation. It is the step before a decision. Suspension requires "adequate evidence" (a standard like probable cause). Debarment requires a preponderance of the evidence, decided by an agency official. Of the 108 referred, there is one completed government-wide debarment: Hafez Mousa Mohammed Mousa, an East Jabaliya Battalion operative who coordinated communications during the October 7 attacks while serving as an UNRWA school principal. This is the one case that has completed the full process, and it was upheld [VERIFIED]. Criminal referrals to the Justice Department are described as forthcoming. None has been announced as filed
- August 2026 — further referrals, and a case the process itself dropped: the OIG referred more staff. They included three UNRWA teachers said to be Hamas or Palestinian Islamic Jihad operatives — one said to have helped abduct hostages on October 7, one to have held civilians hostage inside Gaza, and one, tied to PIJ, said to have crossed into Israel during the attack. They also included Faisal Al-Naami, a UNRWA social worker filmed loading the body of hostage Yonatan Samerano into a vehicle after the abduction from Kibbutz Be'eri (footage released by Israel in 2024; the body recovered by the IDF in 2025) [DOCUMENTED — OIG referrals; Israeli-released footage]. The one-standard caveat here is telling: Al-Naami's debarment was not completed — it was dropped because he is reported dead. That is a concrete reminder. A referral is an allegation at a stage before decision, and not every name on the list leads to an action [DOCUMENTED — the dropped debarment reported by Israeli press]
- The point the process status does NOT mean: "unadjudicated" is not the same as "unconfirmed because UNRWA disagrees." UNRWA's response settles nothing either way (below). The label means only that 107 of 108 referrals await a decision. And every completed review, at every standard, has substantiated cases rather than cleared the list: the UN's own investigation validated 9 of the original 19 within its narrow scope; the US's one finished case was upheld; and three successive investigations ran 12 → 19 → 108, rising each time more investigative capacity was applied [VERIFIED — each figure]
- The evidentiary posture, stated honestly: the OIG says it got its evidence "through independent sources and methods," after the UN's oversight office declined to share its unredacted file. But the OIG does not say how much of it comes from Israeli intelligence. The underlying material is not public, and no wire service has independently examined the list. That absence cuts both ways [VERIFIED — OIG congressional testimony; the absence checked directly]. The nature of that gap is now documented from the US side. A September 2026 USAID OIG review of the UN's OIOS investigation reported "major deficiencies": OIOS declined to assess whether the staff were Hamas members at all, and it left out Israeli intercept intelligence that the OIG said it "found no reason to doubt" [DOCUMENTED — the OIG review as reported by Jerusalem Post; the primary OIG record separately notes OIOS redacted subject names, rendering its report "unusable" to the OIG]. The UN's defensible counter sits on the same record: OIOS required corroboration beyond intelligence it could not check itself and that Israel largely withheld. So the US calls "deficiency" what the UN treats as evidentiary rigor. Both framings are printed here. An earlier leaked account of the case file had suggested the same gap in standards and scope [SINGLE-SOURCE leak — now corroborated by the OIG review]. What the OIG review does not establish is the harder claim that the UN "did not investigate, but protected" UNRWA. OIOS did investigate, and it terminated nine staff. "deficient and shielding key questions" is supported; "deliberate cover-up" is an inference of intent, not a finding [ANALYTICAL FRAMEWORK]
- UNRWA's response, in its own words — and its conduct: six days after the referrals, UNRWA terminated 70 Gaza staff. It said the decision "was taken further to an assessment of the safety and security of UNRWA operations," "is not part of a disciplinary process and does not constitute in any way a validation of the claims," and that Israel has given "no response to date" to UNRWA's repeated requests for the underlying evidence [VERIFIED — UNRWA's June 11, 2026 statement]. Three readings are on the record: a US official called the firings "a direct response" to the OIG investigation; UNRWA calls them a safety measure; and Israeli press reported fear that the accused would be targeted [DOCUMENTED — all three printed]. The evidence-sharing gap is a real process fact, and it is printed. But an agency does not fire 70 employees over allegations it thinks are baseless. Read honestly, UNRWA's conduct points toward the allegations being serious, not against it [ANALYTICAL FRAMEWORK]
- The broader probe — scope, not findings: beyond the 108 referrals, US reporting (July 2026) describes a wider federal examination of ~1,500 UNRWA staff suspected of Hamas or other terror ties. It also describes proposals to strip the agency's immunity or to designate it a foreign terrorist organization [DOCUMENTED as reported — Fox News]. Two cautions travel with this number, and neither can be dropped. It comes from a single major outlet, and it counts the suspected, not the substantiated. It is not interchangeable with the 108 referred or the 9 the UN itself terminated. A count of who is being looked at is not a count of who has been found [ANALYTICAL FRAMEWORK]
- The running ladder, for honest scale: 12 originally alleged (January 2024) → 19 investigated by the UN → 9 terminated on the UN's own standard → 108 referred by the OIG (≈0.8% of ~13,000 Gaza staff) → 70 fired by UNRWA as a safety measure → 1 completed debarment, upheld → 0 criminal charges. Every number is real. Which one measures "infiltration" depends on the standard of proof you apply [VERIFIED — each figure]
What is not established¶
- The systematic claim — that ~10% of Gaza staff, or "significant numbers," are Hamas/Islamic Jihad operatives — has not been ruled on by any court. The independent Colonna Review (April 2024) found UNRWA has "a more developed approach to neutrality than other similar UN or NGO entities." It also found that Israel had not provided evidence for the mass-membership claim [VERIFIED via the Colonna report]. The ICJ's advisory opinion (October 22, 2025) likewise found the broad infiltration allegations were not backed by sufficient evidence on the record then before it, while calling UNRWA's role in Gaza indispensable [VERIFIED via the ICJ advisory opinion; Israel rejects the opinion, which is non-binding]. The OIG's June 2026 referrals are, in effect, the US answer to that finding — two official records now point in opposite directions. One is built on an administrative-evidence standard. The other is built on a judicial-evidence record that predates the referrals. This site reports both, and it labels the question what it is: unadjudicated at scale — with every completed process to date substantiating cases rather than clearing the list [VERIFIED — the framing from the tagged records above]
The deeper problems are structural — by design, not by conspiracy [ANALYTICAL FRAMEWORK]¶
Three of UNRWA's most criticized features would exist even with a perfectly clean staff. They are simply how the agency was built:
- The hereditary mandate: UNRWA registration passes down to descendants with no end — unique among the world's refugee frameworks. So the registered population grew from ~700,000 (1949) to ~5.9 million instead of shrinking. No other refugee population on earth has a matching arrangement. Nor did the ~850,000 Jewish refugees from Arab lands get any agency at all [DOCUMENTED via the institutional record; see the Jewish-refugees FAQ]
- The curriculum: UNRWA schools teach the host authority's textbooks — the Palestinian Authority's — by standard UN design. So documented problems in the textbooks enter through that design choice, not through operatives [DOCUMENTED via curriculum-review reports]
- Local hiring in a Hamas-governed territory: the agency hired ~13,000 staff from a population governed by Hamas for sixteen years. That makes social and political entanglement all but certain — for UNRWA, or for any organization using the same model there [ANALYTICAL FRAMEWORK]
Defenders answer with design arguments too. Host curricula and local hiring are standard practice, they say, and the mandate reflects an unresolved political question rather than creating it. That exchange — about architecture and incentives — is the serious UNRWA debate. It is a different kind of question from the infiltration one.
What has happened as a result¶
The January 2024 allegations triggered ~16 donor suspensions (~$450M). Most donors resumed after the OIOS and Colonna processes. But US funding is now barred by standing statute — the FY2026 appropriations act turned the time-limited prohibition into ongoing law. Israel's Knesset banned UNRWA from Israeli territory and cut off official contact (laws effective January 30, 2025, tightened December 2025). International staff have been barred from Gaza and the West Bank since January 2025. The agency's East Jerusalem headquarters compound was partly demolished in January 2026. As of mid-2026 the agency faces a $220M shortfall. It dismissed ~600 Gaza staff on financial grounds in January 2026 (separate from the June terror-allegation terminations). It has lost 380+ staff killed in the war — the highest UN toll in any conflict. And it is in a leadership transition: Commissioner-General Lazzarini's term ended March 2026, with his warning that the agency "may soon no longer be viable." Christian Saunders serves as acting head [DOCUMENTED via congressional records, Knesset legislation, UN and press reporting].
How to use this honestly¶
"UNRWA is Hamas" claims too much. No process has established organizational direction, and 108 referred is under 1% of the Gaza staff. "The accusations are a smear" is now contradicted by every process that has finished: the UN's own nine terminations, the tunnel under its headquarters, the US's one completed and upheld debarment, and UNRWA's own 70 firings. The statement that survives every fact-check as of mid-2026: individual UNRWA staff participated in terrorism — documented and admitted; a US inspector general has alleged Hamas military affiliation at three-digit scale under an administrative standard, with one case completed and upheld and the rest pending; no judicial body has ruled; and the agency's deepest problems are structural, built into a mandate the UN's member states have renewed for 75 years.
Go deeper: The UN and Israel: The Institutional Record — the mandate-design analysis in full · Jewish refugees from Arab lands · Humanitarian Context · The Gaza genocide question