Genocide & Mass Atrocities: Does 'Never Again' Mean Anything?

The phrase "Never Again" entered the global lexicon after the liberation of Nazi concentration camps in 1945. Eighty years later, it functions less as a commitment than as a recurring epitaph — inscribed after each mass atrocity the international community failed to prevent, then quietly shelved until the next one.

Since the United Nations General Assembly unanimously adopted the Convention on the Prevention and Punishment of the Crime of Genocide in 1948 [1], the world has witnessed mass killings in Cambodia, Rwanda, Bosnia, Darfur, Myanmar, Ethiopia's Tigray region, and now Sudan and Gaza. In nearly every case, credible early warnings existed. In nearly every case, the international response came too late, was too weak, or never arrived.

This report examines why. It traces the institutional failures, political calculations, legal limitations, and structural incentives that have made genocide prevention one of the most comprehensively broken promises in modern history.

Source: United States Holocaust Memorial Museum / Genocide Watch / Academic Consensus Estimates

Data as of Dec 1, 2024CSV

The Legal Record: What Counts as Genocide

Genocide has a precise legal definition under the 1948 Convention: acts committed with the specific intent to destroy, in whole or in part, a national, ethnic, racial, or religious group [1]. That specificity — particularly the requirement to prove "special intent" (dolus specialis) — has made legal findings of genocide rare.

The International Criminal Tribunal for Rwanda (ICTR), established in 1994, produced the first-ever conviction for genocide by an international court when it found Jean-Paul Akayesu guilty in 1998 [2]. The International Criminal Tribunal for the former Yugoslavia (ICTY) ruled that genocide occurred at Srebrenica in July 1995 [3]. The International Court of Justice confirmed in its 2007 Bosnia v. Serbia ruling that the Srebrenica massacre constituted genocide, though it found that Serbia had not committed genocide more broadly — because Bosnia could not prove the "special intent to destroy" beyond the Srebrenica enclave [4].

Beyond these tribunal findings, the ICJ has adjudicated or is currently adjudicating genocide cases involving Myanmar (brought by The Gambia in 2019) and the Gaza Strip (brought by South Africa in 2023) [5]. The ICC has charged Sudanese President Omar al-Bashir with genocide for Darfur since 2009 — the first sitting head of state to face such charges [6].

The gap between events widely recognized as mass atrocities and those legally classified as genocide is vast. Cambodia's Khmer Rouge killed an estimated 1.7 million people between 1975 and 1979; the Khmer Rouge tribunal convicted two leaders of genocide in 2018, but only for crimes against the Vietnamese and Cham Muslim minorities, not the broader Cambodian population [7]. The Armenian Genocide of 1915, which killed an estimated 1.5 million people, has been recognized by over 30 countries but has never been adjudicated by an international court — and Turkey continues to deny it constituted genocide [7].

The practical effect: mass atrocities regularly kill hundreds of thousands before any legal finding occurs, if one ever does. The median time between the onset of mass killing and an international legal determination is measured in years, often decades. International intervention, when it happens at all, typically arrives after the worst violence has already occurred.

The Warning-to-Action Gap: Rwanda, Srebrenica, and Darfur

The distance between what the international community knew and what it did is the central indictment of the post-Holocaust order.

Rwanda: 100 Days, 800,000 Dead

On January 11, 1994, General Roméo Dallaire, commander of the UN Assistance Mission for Rwanda (UNAMIR), sent his now-infamous "Genocide Fax" to UN headquarters in New York. An informant had told Dallaire that Hutu extremists were planning the extermination of Tutsis, and that militia members could "kill up to 1,000 Tutsis in 20 minutes" [8]. Dallaire sent five additional warnings between January 22 and March 13 [8].

The U.S. Central Intelligence Agency, in January 1994, predicted that renewed conflict could kill up to 500,000 people. Belgian intelligence feared a genocide was imminent by February [8]. Human Rights Watch documented illicit arms flows, militia training, and insider accounts of extermination planning months before April [9].

UN headquarters instructed Dallaire not to act on the intelligence and not to seize weapons caches [8]. When the genocide began on April 6, the Security Council's response was to reduce UNAMIR's troop contingent from 2,500 to 270 on April 21 — three weeks into a genocide that would kill 800,000 people in 100 days [10]. Resolution 918, which expanded UNAMIR's mandate and imposed an arms embargo, was not passed until May 17, six weeks into the killing [10]. An independent UN inquiry later concluded that the failure was driven by "a lack of political will" among member states [11].

Srebrenica: Genocide Under UN Protection

The fall of Srebrenica on July 11, 1995 represents an even more direct failure. The town had been declared a UN "safe area" in 1993. Dutch peacekeepers (Dutchbat) were stationed there — under-equipped, under-supplied, and operating without a clearly defined mandate [12].

By early 1995, supply convoys had nearly stopped reaching the enclave. Soldiers patrolled on foot because they had no fuel. Civilians were scavenging garbage dumps for food [12]. When Bosnian Serb forces under General Ratko Mladić launched their offensive, Dutch peacekeepers requested NATO air strikes repeatedly. UN command in Sarajevo denied those requests [13]. Over 8,000 Bosnian Muslim men and boys were systematically executed over the following days.

A seven-year Dutch government investigation concluded that an "anti-intelligence attitude" had prevailed — Dutch military and political leaders believed peacekeeping operations did not require intelligence gathering [12]. The report's publication in 2002 caused the resignation of Prime Minister Wim Kok [12].

Darfur: The First Genocide of the 21st Century

In Darfur, the pattern repeated with additional layers of obstruction. The conflict began in 2003. By 2004, the U.S. State Department had gathered enough evidence for Secretary of State Colin Powell to declare before the Senate Foreign Relations Committee that "genocide has been committed in Darfur" [14]. The ICC issued an arrest warrant for President Bashir in 2009 on charges of genocide, war crimes, and crimes against humanity [6].

Yet the Security Council never authorized forceful intervention. A hybrid AU-UN peacekeeping force (UNAMID) was deployed in 2007 but was chronically underfunded and obstructed by the Sudanese government. An estimated 300,000 or more people died and 2.5 million were displaced [14]. Bashir traveled internationally for years, visiting countries that were parties to the Rome Statute and were legally obligated to arrest him. None did.

The Responsibility to Protect: A Doctrine in Ruins

In 2005, all UN member states unanimously adopted the Responsibility to Protect (R2P) at the World Summit — the principle that sovereignty is not a shield against atrocity crimes, and that the international community has a responsibility to act when a state fails to protect its population from genocide, war crimes, ethnic cleansing, and crimes against humanity [15].

Twenty years later, the UN Secretary-General's own 2025 report on R2P acknowledged that atrocity crimes have increased globally since its adoption, "exposing a persistent gap between promises and meaningful action" [15].

The Libya Problem

Libya in 2011 is the only case where the Security Council achieved consensus to authorize military intervention under R2P's "Pillar 3" — the use of coercive measures including force [16]. Resolution 1973 authorized member states to take "all necessary measures" to protect civilians. The subsequent NATO air campaign helped topple Muammar Gaddafi's regime.

What followed has haunted R2P ever since. Libya descended into civil war, state collapse, and became a hub for human trafficking and militant activity [17]. Russia and China — and many states in the Global South — accused NATO of misusing the civilian protection mandate to pursue regime change. This perceived betrayal of the resolution's terms has been cited by Russia and China as justification for vetoing subsequent action, most notably on Syria [16].

Syria: The Veto as Weapon

Between 2011 and 2024, Russia vetoed at least 16 Security Council resolutions on Syria, often with Chinese support [16]. The Syrian civil war killed an estimated 500,000 people and displaced half the country's population. Chemical weapons were used against civilians multiple times. No R2P intervention was authorized.

The defenders of the vetoes argued that Western interventions consistently produced worse outcomes than the conflicts they aimed to resolve — pointing to Libya's post-intervention chaos, Iraq's destabilization, and Afghanistan's ultimate collapse. Russian officials explicitly cited the Libya precedent as proof that humanitarian mandates serve as pretexts for regime change [16].

The Veto Scorecard

Since 2011, the veto has been used dozens of times by P5 members to block resolutions addressing atrocity situations in Syria, Israel and the Occupied Palestinian Territory, Sudan, South Sudan, Yemen, Mali, and Ukraine [15]. Reform efforts — including the ACT Group's Code of Conduct urging veto restraint in atrocity situations, the French-Mexican veto restraint initiative, and General Assembly Resolution 76/262 requiring a General Assembly debate after every veto — have produced no structural change [15].

The Funding Gap: Prevention vs. Response

The disparity between what the international system spends preventing mass atrocities and what it spends responding to them is staggering.

The UN approved $5.6 billion for 14 peacekeeping operations for 2024-2025 [18]. The UN Office on Genocide Prevention and the Responsibility to Protect — the institution specifically tasked with early warning and prevention — operates with a staff and budget that a 2024 Stimson Center assessment described as inadequate for its mandate [19]. Exact figures are difficult to isolate due to changing budget categories, but estimates place the Office's annual budget in the low single-digit millions — less than one-tenth of one percent of peacekeeping expenditure [19].

The asymmetry extends to personnel. Peacekeeping operations deploy over 70,000 uniformed personnel globally [18]. The genocide prevention office operates with a small team that cannot maintain sustained engagement with the dozens of at-risk situations identified by its own monitoring [19].

The practical consequence: the international system is structured to respond to mass atrocities after they occur, not to prevent them. A 2024 assessment found that a minority of situations identified by the Office on Genocide Prevention as at-risk received preventive diplomatic or economic measures before violence escalated [19].

For context on resource allocation priorities: as of 2024, military expenditure as a percentage of GDP stood at 3.4 percent for the United States, 2.3 percent for the United Kingdom and India, 1.9 percent for Germany, and 1.7 percent for China [33]. The contrast between national military budgets measured in hundreds of billions and the genocide prevention office's budget measured in single-digit millions illustrates where institutional priorities lie.

Source: World Bank Open Data

Data as of Dec 31, 2024CSV

Gaza: The Most Contested Case in a Generation

No current situation illustrates the contested nature of genocide determination more sharply than Gaza. The debate is substantive, involves serious legal and moral arguments on both sides, and has divided governments, legal scholars, and human rights organizations.

The Case That Israel Is Committing Genocide

South Africa filed its case against Israel at the ICJ on December 29, 2023. On January 26, 2024, the Court ruled that South Africa's claims were plausible and ordered provisional measures directing Israel to prevent acts within the scope of the Genocide Convention, ensure its military did not commit such acts, prevent incitement to genocide, and enable humanitarian assistance [20]. On March 28, 2024, the Court ordered Israel to ensure basic food supplies, finding that famine conditions existed. On May 24, 2024, by a vote of 13-2, the Court ordered a halt to Israel's Rafah offensive [20].

Proponents of the genocide characterization cite the scale of destruction: by the end of 2024, an estimated 78,318 people had been killed [21]. A Max Planck Institute for Demographic Research study estimated total violent deaths between 100,000 and 126,000 by October 2025 [21]. Israeli military intelligence's own data, as reported in May 2025, listed 8,900 Hamas and Palestinian Islamic Jihad fighters as confirmed or probably dead — meaning named combatants accounted for roughly 17 percent of casualties, with civilians constituting approximately 83 percent [22]. Children accounted for at least 30 percent of total casualties [22].

Scholars and institutions have characterized the situation in terms of genocide. Raz Segal, an Israeli Holocaust and genocide studies scholar at Stockton University, called it "a textbook case of genocide" in October 2023 [23]. The Lemkin Institute for Genocide Prevention issued a genocide alert. Over 800 scholars of international law and genocide studies signed statements characterizing Israeli actions as constituting genocide or creating imminent risk of genocide [23]. Critics of Israel cite statements by Israeli officials as evidence of genocidal intent: President Isaac Herzog stating "it is an entire nation that is responsible," Defense Minister Yoav Gallant's reference to "human animals," Finance Minister Bezalel Smotrich's advocacy for "voluntary migration," and Heritage Minister Amichai Eliyahu's suggestion of using a nuclear weapon on Gaza [23].

The systematic destruction of hospitals, universities, mosques, churches, and cultural heritage sites; the forced displacement of 1.9 million people (approximately 85 percent of Gaza's population); and the obstruction of humanitarian aid — including the defunding of UNRWA based on Israeli claims that were not independently verified at the time — form the evidentiary basis for the genocide allegation [23].

The Case That Israel Is Not Committing Genocide

Genocide under international law requires specific intent (dolus specialis) to destroy a group "as such" — not merely the infliction of mass casualties during armed conflict [1]. Legal scholars at the Jerusalem Center for Public Affairs and the Kohelet Policy Forum argue that Israel's military operations target Hamas's command infrastructure, not Palestinians as a people [24].

The civilian casualty ratio, while devastating, is comparable to or lower than other cases of urban warfare — Mosul, Raqqa, and Fallujah — where no genocide allegations were made [24]. Hamas's documented strategy of embedding military infrastructure within civilian areas — including hospitals, schools, and tunnel networks beneath residential neighborhoods — means that civilian casualties, however tragic, are a foreseeable consequence of urban combat, not evidence of genocidal intent [24].

Israel dropped leaflets, made phone calls, sent text messages, and issued evacuation warnings before major operations — actions inconsistent with a campaign designed to maximize civilian deaths [24]. The Gaza Health Ministry, which generates the casualty figures cited by international organizations, is operated by Hamas and does not systematically distinguish combatants from civilians in its reporting, which complicates independent verification [24].

On the ICJ proceedings, defenders of Israel note that the January 2024 provisional measures order was procedural — the Court found South Africa's claims "plausible" (meeting the low threshold for provisional measures) but made no finding on whether genocide has occurred [24]. Israel submitted its counter-memorial to the ICJ on March 14, 2026, and the merits phase will determine the substantive legal question [20].

The broader concern: if every war with high civilian casualties meets the threshold for genocide, the term loses its legal precision and moral weight.

The Double Standard in Both Directions

Critics of Israel point to a stark contrast in international response. Russia's invasion of Ukraine prompted swift sanctions, weapons deliveries to Ukraine, and ICC arrest warrants for Putin [25]. The response to Palestinian suffering included the United States vetoing multiple UN Security Council ceasefire resolutions and continuing arms sales to Israel [25].

Critics of the genocide framing point to an equally uncomfortable asymmetry in public attention. The Tigray war in Ethiopia killed an estimated 300,000 to 800,000 people between 2020 and 2022 — potentially the deadliest conflict of the 21st century — and generated a fraction of the global protest and media coverage directed at Gaza [26]. The Uyghur internment in Xinjiang — with approximately 500,000 in detention, forced sterilization campaigns, and what the U.S., U.K., and Canadian parliaments declared genocide — has not produced sanctions comparable to those imposed on Russia [27]. The U.S.-armed Saudi coalition killed tens of thousands of civilians in Yemen with no genocide allegations [25]. Sudan's ongoing civil war, with an estimated 400,000 dead and 11 million displaced, receives marginal international attention relative to its scale [28].

Why does Gaza generate global protests while 600,000 dead in Tigray barely made the news? The answers are multiple and contested: the role of social media in amplifying some conflicts over others; the degree to which Western governments are directly implicated through arms sales; longstanding engagement with the Israeli-Palestinian conflict in both Western and Arab political discourse; and, as some critics allege, antisemitism operating under the banner of human rights. Each explanation has evidence supporting it; none is sufficient alone.

The ICC's Credibility Crisis

The International Criminal Court, established by the Rome Statute in 2002, was designed to end impunity for the worst crimes. Its record suggests otherwise.

As of 2025, the ICC has issued arrest warrants for multiple current and former heads of state: Omar al-Bashir (Sudan, 2009), Muammar Gaddafi (Libya, 2011, mooted by his death), Uhuru Kenyatta (Kenya, charges withdrawn), Vladimir Putin (Russia, 2023), and Benjamin Netanyahu (Israel, 2024) [6]. Of sitting leaders, only Rodrigo Duterte of the Philippines has been arrested and transferred to The Hague — and that occurred in March 2025, years after he left office [6].

Bashir traveled to ICC member states for over a decade without being arrested. Putin continues to travel internationally. Netanyahu continues to serve as Prime Minister. The ICC has no police force and depends entirely on member state cooperation for enforcement — cooperation that has been inconsistent at best [6].

The court has secured convictions overwhelmingly against African defendants, prompting accusations of selective justice. Several African states have threatened or initiated withdrawal from the Rome Statute. The United States, China, Russia, India, and Israel are not members [6].

The structural problem: the ICC issues warrants it cannot enforce against leaders whose states or allies can shield them from accountability. Its deterrent effect is, at minimum, unproven.

The Current Risk Map

The Early Warning Project, a joint initiative of the Simon-Skjodt Center for the Prevention of Genocide at the United States Holocaust Memorial Museum and Dartmouth College's Dickey Center, produces annual risk assessments for intrastate mass killing across more than 160 countries [29].

For 2025-2026, only Myanmar has a two-year estimated risk exceeding 10 percent [29]. The model indicates that in any given year, approximately two out of every three countries that later experienced a new mass killing had ranked among the top 30 in the assessment [29].

The Global Centre for the Responsibility to Protect's July 2025 populations-at-risk assessment identified 20 situations of concern: Afghanistan, Belarus, Cameroon, Central African Republic, the Central Sahel (Burkina Faso, Mali, Niger), China, North Korea, the Democratic Republic of Congo, Ethiopia, Haiti, Israel and the Occupied Palestinian Territory, Myanmar, Nicaragua, Nigeria, South Sudan, Sudan, Syria, Ukraine, Venezuela, and Yemen [30].

Sudan: The Catastrophe Hiding in Plain Sight

Sudan's civil war, which began in April 2023 between the Sudanese Armed Forces (SAF) and the Rapid Support Forces (RSF), has produced what may be the world's worst ongoing humanitarian crisis with the least international attention relative to its severity.

The former U.S. envoy for Sudan estimated that as many as 400,000 people had been killed [28]. Between January and June 2025 alone, UN Human Rights documented the deaths of at least 3,384 civilians, nearly 80 percent in Darfur [31]. More than 11 million people have been displaced — the largest displacement crisis in the world [28].

The RSF has carried out systematic, ethnically targeted massacres. In November 2023, RSF forces killed more than 800 people in a multi-day attack on Ardamata in West Darfur, targeting the Masalit ethnic group [28]. In December 2025, a Yale School of Public Health investigation assessed with "high confidence" that the RSF conducted "widespread and systematic mass killings" after seizing El Fasher, including a campaign to conceal evidence through the burial, burning, and removal of remains [32].

Sexual violence has been systematic, with women and girls as young as 14 documented as victims [31].

The global refugee crisis reflects these ongoing atrocities. As of 2025, Syria produces the most refugees globally (5.5 million), followed by Ukraine (5.3 million), Afghanistan (4.8 million), Sudan (2.5 million), and South Sudan (2.4 million) [34]. Sudan's displacement of 11 million people internally, combined with 2.5 million registered refugees, confirms its status as the largest displacement crisis in the world.

Source: UNHCR Population Data

Data as of Dec 31, 2025CSV

What Has Changed Since 1994 — And What Hasn't

The institutional landscape for genocide prevention has changed significantly since the failures of Rwanda and Srebrenica:

New mechanisms: The ICC (2002), the UN Office on Genocide Prevention (2004), the R2P doctrine (2005), the Early Warning Project, the Global Centre for the Responsibility to Protect, and General Assembly Resolution 76/262 on veto accountability all represent institutional innovations that did not exist in 1994 [15].

New legal precedents: The ICTR and ICTY established that genocide, crimes against humanity, and command responsibility could be prosecuted internationally. The ICC has issued warrants against sitting heads of state from multiple continents [6].

Unchanged structural constraints: The Security Council veto remains intact. No enforcement mechanism exists for ICJ or ICC orders. Peacekeeping forces still cannot be deployed without Security Council authorization. The states with the most military capacity to intervene — the P5 — are also the states most frequently implicated in or allied with perpetrators of mass atrocities [15].

Would the institutional changes produce different outcomes if Rwanda or Srebrenica happened today? The evidence from recent cases suggests not. The early warning systems identified risks in Tigray, Myanmar, Sudan, and the Sahel — but identification did not translate into prevention [29]. ICC warrants for Bashir, Putin, and Netanyahu have not constrained their behavior. The Security Council remains paralyzed by vetoes on the most politically charged situations.

Beyond Military Force: What the Evidence Shows

The debate over intervention often defaults to a binary — military force versus inaction — that obscures the range of tools available and their track record.

Targeted sanctions: Economic sanctions targeting specific individuals and entities have been applied in Myanmar, Sudan, Russia, and other cases. Their impact on reducing violence is mixed. Sanctions on Myanmar's military junta after the 2021 coup have not reversed the coup or stopped atrocities against ethnic minorities. Sanctions on Russian oligarchs and institutions after the Ukraine invasion have imposed economic costs but have not ended the war [25].

ICC indictments: The deterrent value of international criminal accountability remains debated. Proponents argue that warrants signal that impunity has costs and can constrain travel and diplomatic engagement. Critics note that Bashir evaded arrest for over a decade, Putin has faced no operational constraints from his warrant, and the ICC's record of convictions is thin relative to its mandate [6].

Diplomatic pressure and mediation: The Cessation of Hostilities Agreement that ended the Tigray war in November 2022 was brokered by the African Union, not the UN or Western powers [26]. Diplomatic engagement can reduce violence, but its effectiveness depends on the parties' incentives — and external mediators have limited ability to alter those incentives without economic or military backing.

The intervention vs. non-intervention ledger: Critics of military intervention point to Libya's post-2011 collapse and Iraq's destabilization as evidence that external force produces state failure and long-term instability. Defenders of intervention point to Rwanda and Srebrenica, where non-intervention produced genocide. The Kosovo intervention in 1999 — conducted without Security Council authorization — is often cited as a case where military action prevented a larger atrocity, though it too was followed by ethnic violence against Serbs [17].

The honest accounting: both intervention and non-intervention have produced catastrophic outcomes depending on context. Neither a blanket doctrine of sovereignty nor a blanket doctrine of humanitarian intervention fits the evidence. The determining factors appear to be the specificity of the military objective, the availability of a viable political settlement, and the commitment to post-conflict stabilization — none of which can be guaranteed in advance.

The Structural Problem

The architecture of international genocide prevention rests on a contradiction: the states responsible for authorizing collective action are the same states whose geopolitical interests most frequently obstruct it. The P5 veto ensures that atrocities committed by permanent members (China's treatment of Uyghurs, Russia's actions in Ukraine) or their allies (Saudi Arabia in Yemen, Israel in Gaza) cannot trigger mandatory Security Council action.

The UN genocide prevention office is funded at a fraction of a percent of peacekeeping expenditure. Early warning systems produce assessments that generate reports but not action. The ICC issues warrants it cannot enforce. R2P was adopted unanimously and has been applied once — in a case that destroyed the doctrine's credibility.

The system is not failing to work as designed. It is working exactly as designed — by states that built in escape hatches for themselves and their allies while creating the appearance of a rules-based order.

Whether "Never Again" can ever mean what it says depends on whether the international community is willing to restructure the incentives that make inaction the default. The 2025 Secretary-General's report on R2P acknowledged the gap between promise and performance [15]. It proposed no mechanism to close it.

The dead in Rwanda, Srebrenica, Darfur, Tigray, Myanmar, and Sudan testify to a consistent pattern: the international community possesses the information, the legal frameworks, and the military capacity to prevent genocide. What it lacks is the political will to use them — and a system designed to produce that will rather than obstruct it.