The International Court of Justice issued one of the most closely watched legal decisions of the Gaza war on Friday. South Africa has accused Israel of violating the 1948 Genocide Convention through its military campaign in Gaza and asked the Court to impose emergency measures while the full case proceeds. The Court did not decide whether genocide has occurred. That question belongs to the merits of the case and could take years to resolve. It also did not grant South Africa’s request for an order requiring Israel immediately to suspend its military operations. Instead, the Court concluded that it had prima facie jurisdiction and that at least some of the rights South Africa claims under the Convention are plausible enough to require protection while proceedings continue.
That distinction is essential because the word genocide carries extraordinary legal and moral weight, and Friday’s ruling is already being interpreted politically in ways that go beyond what the Court actually decided. Saying that the ICJ has ruled Israel guilty of genocide would be false. Saying that the Court dismissed South Africa’s case would also be false. The judges refused Israel’s request to remove the case at this stage and imposed several legally binding provisional measures while leaving the ultimate allegation unresolved.
The measures themselves are substantial. By 15 votes to two, the Court ordered Israel, in accordance with its obligations under the Genocide Convention, to take all measures within its power to prevent acts covered by Article II of the Convention against Palestinians in Gaza and to ensure its military does not commit those acts. By 16 votes to one, it ordered Israel to prevent and punish direct and public incitement to genocide and to take immediate and effective measures enabling urgently needed basic services and humanitarian assistance. Israel must also preserve relevant evidence and submit a report to the Court within one month explaining the measures it has taken.
The Court nevertheless declined to order the immediate suspension of military operations that South Africa requested. That matters because Israel argues that it is exercising its right to defend itself against Hamas following the 7 October attacks, while South Africa argues that the scale and character of Israeli operations and conditions imposed on Gaza engage Israel’s obligations under the Genocide Convention. The provisional ruling attempts to protect rights potentially at risk without deciding the ultimate factual and legal dispute before the evidence has been examined fully.
This is what provisional measures are designed to do. Courts often confront situations where waiting for a final judgement could make the judgement meaningless. If a disputed building is demolished while litigation continues, determining five years later that demolition was unlawful cannot reconstruct the building automatically. International litigation involving human life presents the same problem at an incomparably greater scale. The Court therefore asks whether rights requiring protection are sufficiently plausible, whether there is urgency and whether irreparable prejudice could occur before the final judgement.
This means the evidentiary threshold at this stage is deliberately different from the threshold required eventually to establish genocide. Genocide is among the most difficult international crimes to prove because it requires more than demonstrating enormous civilian suffering or even unlawful killing. The Convention requires prohibited acts committed with the specific intent to destroy, in whole or in part, a national, ethnical, racial or religious group as such. Mass civilian death can constitute war crimes or crimes against humanity without satisfying that specific genocidal-intent requirement.
The distinction is morally uncomfortable because legal categories can sound like rankings of suffering. They are not. Saying conduct does not meet the legal definition of genocide does not mean the conduct is acceptable. International law contains several categories because different prohibited acts require different evidence and create different legal responsibilities. Precision matters precisely because accusations are serious.
South Africa therefore faces a much more demanding task during the eventual merits stage than it faced in obtaining provisional measures. It will need to establish not simply catastrophic conditions in Gaza but the legal elements required under the Convention and Israel’s responsibility for them. Israel will have the opportunity to contest evidence, provide military and humanitarian explanations for its conduct and argue that its objective is defeating Hamas rather than destroying Palestinians as a group.
That process is what courts are for.
The purpose of adjudication is not confirming whichever conclusion people already reached politically.
This is especially important because the Gaza war has produced an extraordinarily polarised information environment. One side can cite the devastation of Gaza, displacement, restrictions on humanitarian supplies and statements by Israeli officials as evidence supporting South Africa’s allegations. The other can cite Israel’s stated military objectives, evacuation warnings, humanitarian measures and Hamas’s operations from civilian areas as evidence against genocidal intent.
A court needs to distinguish evidence from narrative.
That takes time.
Social media does not.
The difference between legal time and political time is therefore enormous. Millions of people have already decided whether the word genocide applies. The ICJ may take years to produce a final judgement. By then, governments will have changed, military operations may have ended and the political environment could look completely different.
This raises an obvious question: what is the value of international law if final judgements arrive after the events they are supposed to regulate?
Provisional measures provide one answer.
They allow courts to intervene legally before deciding ultimate responsibility.
But the deeper answer concerns what international courts can and cannot do.
The ICJ does not possess an international police force. It cannot send officers into Israel to enforce its order or physically compel humanitarian deliveries. Its orders are binding, but implementation ultimately depends substantially on states, international organisations and political pressure.
Critics therefore argue that international law is weak because powerful states obey only when convenient.
There is truth in that criticism.
International law lacks the centralised enforcement machinery domestic law normally possesses.
But concluding that absence of perfect enforcement means absence of power would be a mistake.
Domestic law itself is not obeyed only because police stand beside every citizen. Most legal systems function because law changes incentives, expectations and legitimacy. Businesses follow contracts because courts exist in the background. Governments comply with judicial decisions partly because non-compliance creates institutional and political costs.
International law works through a weaker version of the same mechanism.
A judgement changes the political price of behaviour.
That can matter particularly for Israel because it depends heavily on relationships with the United States and European governments. If allied governments conclude that particular Israeli policies create serious risks under international law, they face their own legal and political questions regarding military assistance, diplomatic support and arms exports.
The Court therefore does not need the capacity to enforce its judgement alone if its decisions influence actors who possess other forms of leverage.
This is distributed enforcement.
Courts establish legal findings.
Governments control diplomacy and sanctions.
Domestic courts can examine national legal obligations.
Civil society generates political pressure.
Companies assess reputational and legal risks.
No institution controls the entire enforcement chain.
Together they can change incentives.
Whether that happens depends heavily on politics.
This is why the accusation that international law is “political” needs clarification. Law and politics are not opposites in international relations because implementation necessarily occurs through states. The more useful question is whether legal standards constrain political decisions or merely provide language governments use selectively against adversaries.
The Gaza case is therefore a major credibility test for Western governments.
The United States and European states have repeatedly invoked international law in condemning Russia’s invasion of Ukraine. They have argued that sovereignty, territorial integrity and civilian protection are principles rather than preferences applied only when strategically convenient.
Israel is a close Western partner.
That makes consistency harder.
It also makes consistency more valuable.
If international humanitarian and genocide-law obligations are interpreted rigorously when Russia is accused and dismissed reflexively when Israel is accused, governments across the Global South will reasonably conclude that the “rules-based international order” contains different standards depending on alignment.
The opposite error would be treating Israel as presumptively guilty because applying pressure to an ally demonstrates consistency.
Consistency does not mean reaching identical conclusions in different cases.
It means applying the same evidentiary seriousness and legal principles regardless of which state is accused.
This is precisely why Friday’s ruling should not be exaggerated.
People sympathetic to Palestinians weaken the legal argument when they describe the Court as having already found genocide.
People sympathetic to Israel weaken their argument when they describe the case as though the Court found nothing serious enough to examine.
The actual ruling is more complicated and therefore more useful.
The Court found enough risk under the Genocide Convention to impose binding preventive obligations while refusing to decide the ultimate allegation prematurely.
That is what judicial restraint looks like.
The voting pattern is also interesting. Several measures passed by overwhelming majorities, including the requirement to facilitate humanitarian assistance, which passed 16–1. Even Aharon Barak, the Israeli judge ad hoc appointed for the case, supported measures concerning humanitarian assistance and preventing and punishing incitement to genocide, although he opposed several others.
That should make it harder to dismiss the entire order simply as geopolitical hostility towards Israel.
At the same time, disagreement among judges demonstrates that international law does not eliminate interpretation.
Judges evaluate evidence and legal thresholds differently.
Law structures disagreement.
It does not abolish it.
The humanitarian-assistance order may become the most immediately consequential part of the decision. Gaza’s civilian population faces severe shortages of food, water, medicine and functioning infrastructure after months of war. The Court ordered Israel to take immediate and effective measures enabling urgently needed basic services and humanitarian assistance.
This creates a measurable standard against which conduct can be evaluated.
Are more aid deliveries entering?
Are restrictions being reduced?
Can humanitarian organisations distribute supplies?
Are essential services becoming more accessible?
Legal orders become meaningful when they create observable obligations rather than simply expressing concern.
The requirement concerning incitement is equally significant. Genocide law does not regulate only physical acts. Direct and public incitement to commit genocide is itself prohibited under the Convention. The Court therefore ordered Israel to prevent and punish such incitement concerning Palestinians in Gaza.
This matters because wartime rhetoric can change the environment in which military conduct occurs. Dehumanising entire populations makes collective punishment easier politically. Leaders may use extreme language to mobilise frightened or angry publics, but language from officials can influence soldiers, institutions and public expectations.
Speech by people holding state power is not merely speech.
It can become part of the institutional environment surrounding violence.
The principle should apply equally to rhetoric calling for the destruction of Israelis or Jews. Hamas’s founding ideology, statements by militant leaders and the deliberate targeting of civilians deserve equivalent scrutiny under relevant legal standards.
Again, protecting one population does not require minimising threats towards another.
International law becomes credible when civilian protection is non-zero-sum.
The political debate frequently behaves as though recognising Palestinian rights reduces recognition of Israeli security, or recognising Israeli security reduces Palestinian rights.
Law attempts to establish minimum obligations that survive precisely when political empathy becomes selective.
That is why the Genocide Convention exists.
It was created after the Holocaust because the international community recognised that sovereignty cannot provide unlimited protection for governments committing certain acts against populations under their control.
The Convention therefore places duties not only on states accused of genocide but on all parties to prevent and punish genocide.
That creates another difficult question for countries supplying weapons or diplomatic support to parties in conflicts where genocide allegations arise.
How much knowledge creates responsibility?
What preventive actions are reasonable?
When does ordinary military assistance become legally problematic?
Those questions will become increasingly important if the case continues.
They demonstrate how litigation against one state can influence the behaviour of many others.
International courts therefore exercise network power.
The immediate order is directed towards Israel.
The political and legal effects travel through governments, corporations, domestic courts and international organisations connected to Israel.
This resembles economic sanctions in reverse: rather than directly imposing material costs, legal findings alter the risk calculations of actors capable of imposing or withholding them.
The strength of the mechanism depends on the Court’s credibility.
That credibility depends on courts resisting pressure to become extensions of political coalitions.
South Africa itself has geopolitical interests and political motivations.
That does not invalidate its legal case.
Israel has national-security interests and political motivations.
That does not invalidate its legal defence.
Courts exist because the motives of litigants do not determine the truth of their claims.
Evidence does.
This is an important principle beyond Gaza because international litigation is increasingly becoming another arena of geopolitical competition. States use courts strategically because legal legitimacy matters.
Ukraine has brought cases against Russia.
The Gambia brought a genocide case concerning Myanmar.
South Africa has brought this case against Israel.
Legal institutions therefore operate inside geopolitics while attempting to apply standards that transcend it.
Their authority depends on maintaining that distinction convincingly.
The ICJ also faces a fundamental enforcement paradox. If it issues only decisions powerful states are likely to obey voluntarily, it preserves compliance but becomes legally timid. If it issues demanding orders that states ignore, it risks exposing its lack of coercive capacity.
Courts therefore need enough independence to issue law regardless of political convenience while relying on the broader international system to make non-compliance costly.
This is why international institutions cannot be evaluated individually.
The Court needs governments.
Governments need legitimacy.
Civil society needs legal standards.
Domestic institutions translate international obligations into national policy.
The system is decentralised.
Its weakness is obvious.
Its strength is that enforcement can emerge from several directions rather than one global sovereign.
Friday’s ruling will therefore matter partly through what happens next.
Israel has been ordered to report within one month on measures taken to comply. South Africa will presumably scrutinise that report. Other governments will evaluate Israeli conduct. Humanitarian organisations will document conditions in Gaza.
The Court has created a legal benchmark.
Whether that benchmark changes reality depends on whether other institutions use it.
This is why neither triumphalism nor dismissal makes sense.
South Africa has not won the genocide case.
Israel has not been cleared.
The Court has not ordered an immediate ceasefire.
It has not concluded that nothing requiring urgent legal protection is happening.
All four statements can be true simultaneously.
That complexity is precisely why judicial language matters.
Political movements want binary verdicts.
Courts often produce conditional ones.
The eventual merits case will ask whether Israel violated obligations under the Genocide Convention.
Friday’s order asks a different question: whether rights protected by that Convention face sufficient plausible and urgent risk that measures are required before the final judgement.
The Court answered yes.
That answer creates obligations now even while guilt remains unresolved.
This is the broader value of law under uncertainty.
Governments frequently need to act before every fact is known.
The precautionary principle exists in several legal contexts because waiting for certainty can make prevention impossible.
Provisional measures apply the same logic judicially.
Protect potentially endangered rights now.
Determine final responsibility later.
That inevitably creates the possibility that severe restrictions are imposed before ultimate guilt is established.
The alternative creates the possibility that courts discover years later that irreversible harm occurred while they waited.
Law therefore manages risk as well as guilt.
This is why the ICJ’s intervention should be evaluated neither as a conviction nor as meaningless symbolism.
It is an attempt to place legal constraints around an ongoing political and military process before the Court knows enough to resolve the entire dispute.
Whether those constraints hold will tell us something larger about the international system.
The world’s highest court has spoken.
It cannot enforce its words with soldiers.
That does not mean the words have no power.
Their power depends on whether governments that claim to believe in international law treat legal constraints as real when those constraints become politically inconvenient.
That is the actual test beginning now.
