If liberal interventionism was crippled in Iraq, it died in Libya.
Iraq showed the world how easily liberal language could be attached to regime change, then Libya showed that even a more recognisably humanitarian intervention could still end in collapse.
This is not to say the instinct behind liberal interventionism was wrong. The principle behind it is one of the most attractive features of liberal theory: sovereignty cannot be a license for mass slaughter, and dissidents should be protected from tyrants. A liberalism that shrugs at torture, ethnic cleansing, or mass repression because it happens behind another state’s borders is not obviously more principled than one too quick to intervene. It may merely be less demanding.
The central problem for liberal interventionism is that it tries to act on two principles which cannot both be fully satisfied in an anarchic international system. Cosmopolitan liberalism says that individuals possess fundamental rights as human beings, not merely as citizens of a particular state. Why should a dictator be allowed to murder his own people simply because they are “his” people? Simultaneously, political self-determination says that peoples should govern themselves without external coercion. Why should a foreign state decide when another people are no longer governing themselves?
Both of these are liberal and necessary. Yet, they often pull in opposite directions. This is the intervention paradox: liberalism wants to protect individuals from tyrants, but it also wants people to rule themselves.
Locke gives liberal theorists one answer. When rulers violate the people’s rights and reduce them to “slavery under arbitrary power”, they put themselves into a “state of war with the people”, who are then “absolved from any farther obedience”. Political power is not a blank cheque. It is conditional on the purposes for which it was granted: the protection of life, liberty, and property. When a government becomes the organised enemy of those rights, its claim to authority collapses.
Locke’s theory is compelling as it concerns domestic political authority, but when extended to foreign intervention, it is incomplete. If the people rise up and overthrow their tyrannical government, the answer is simple enough. The harder cases are those where repression works and a domestic revolt is unlikely despite reprehensible atrocities. Justifying external involvement is more strenuous because sovereignty protects states from external interference. Who outside that political community has the right to enforce that judgment? How many victims must a government create before that protection expires: ten thousand? A hundred thousand? An entire ethnic group?
The domestic answer to this problem is Weber’s ‘monopoly on legitimate violence’. If you know your neighbour is abusing their child, you face a moral conflict. You want to protect the child, but you don’t want every citizen to become judge, jury and executioner. So, societies created institutions that could legitimately intervene: the police, courts, and government. This is not to say domestic justice is perfect, but it gives force a degree of legitimacy.
In the aftermath of Rwanda, Srebrenica, Kosovo, and the wider post-Cold War debate over humanitarian intervention, the Responsibility to Protect (R2P) doctrine attempted to create a legitimate solution. Adopted at the 2005 UN World Summit, R2P reframed sovereignty as a duty of states to protect their populations from genocide, war crimes, ethnic cleansing, and crimes against humanity. The international community has a duty to help them do so, and if a state is manifestly failing to protect its people, the international community must be prepared to take collective action.
R2P was the most serious institutional attempt to reconcile human rights and sovereignty, setting clear atrocity thresholds and insisting on collective action with international consensus so that intervention wouldn’t become imperialism. Yet, the doctrine relied on exactly what an international system lacks: an impartial and effective enforcement agent. There is no global police force with a settled monopoly on legitimate violence.
In theory, the United Nations is the closest thing we have to a collective arbiter of international legitimacy. In practice, however, the Security Council’s permanent-member veto often produces great-power deadlock. States may agree that atrocities are occurring while disagreeing over who should bear the costs, command the operation, and what political settlement should follow.
In practice, powerful states—especially the United States in the post-Cold War settlement—often acted in its place. But this created selectivity, bias, and legitimacy problems. The United States intervened in Libya, but not in Syria. Too often, the United States seemed to intervene when it allied with other, more strategic interests, and dragged its feet when facing a strategic ally, hurting any claims to legitimacy.
Any serious defence of intervention has to clear three hurdles: motive, authority, and consequence.
First, how do we know it is actually about human rights?
This is not an unfair question. Humanitarian language has often mingled with geopolitical interest. The United States and the Soviet Union both supported coups, armed dictatorships, and maintained alliances with authoritarian regimes when doing so suited their strategic purposes. This does not mean that every intervention is secretly imperialist. But it does mean that power rarely acts without interests, and moral language can be used to launder them. Any intervener must prove a truly humanitarian intent.
Second, authority: who decides when a people is no longer self-determining? It is profoundly dangerous to put the power of revolution and self-determination into the hands of an arbitrary foreign power. However oppressed a population may be, if an external state claims the right to speak for them, it risks substituting its own judgment for theirs. Intervention may be justified only when domestic legitimacy has collapsed more concretely: the regime rules through mass violence, blocks meaningful avenues of political change, and treats basic opposition as an enemy to be destroyed. Even then, the act of intervention itself can obscure whether legitimacy collapsed internally or was destroyed externally.
Third, consequences: will intervention actually improve anything? Violating sovereignty is costly, so there is a burden of proof on the intervener to show that they are the best-positioned actor to act and that the results will be positive. Iraq removed a tyrant and produced occupation, insurgency, sectarian civil war, and regional destabilisation. Libya was justified as preventing an imminent massacre, but ended in collapse and militia rule.
Intervention can destabilise entire regions, triggering proxy wars and state collapse. External involvement may also delegitimise domestic movements and allow regimes to paint protestors as foreign agents. In the end, intervention may make the people it intends to save less free, less safe, and less capable of determining their own future.
Liberal intervention is only justified when all three of these conditions are met: the regime has forfeited the normal protection of sovereignty through grave and systematic abuse; the intervener can make a credible claim to authority that does not simply amount to power appointing itself judge; and there is a serious reason to believe intervention will improve rather than worsen the position of those it claims to protect.
This is the closest liberal interventionism gets to an answer. It cannot eliminate the paradox through clean answers, but it can impose a burden of proof and some necessary conditions. Without the first, intervention is aggression. Without the second, it is domination. Without the third, it is recklessness.
Iran shows how this works in practice. The Iranian regime is plainly a repressive, theocratic dictatorship. The “Woman, Life, Freedom” protests following Mahsa Amini’s death were met with mass arrests, killings, torture, executions, and sexual violence; UN investigators have found that parts of the repression amounted to crimes against humanity.
A self-respecting liberal cannot look at this and say ‘it’s their business’. The Iranian regime has clearly violated the trust on which political authority depends, as the January 2026 protests strikingly displayed. The first condition has been met.
But that does not yet make outside military intervention legitimate. The UN’s veto problem leaves the United States as the most plausible external military actor, but this immediately creates a legitimacy problem: America has a history of intervention in Iranian politics, and Donald Trump makes any ‘humanitarian’ justification suspect.
Even if some Iranian dissidents welcomed external pressure, the regime already frames opposition as foreign-backed; intervention would make that claim easier to sell. It is very doubtful that the second condition is met.
The practical case is no easier. Iran is not Libya. It is large, nationally conscious, and militarily capable. The regime’s display of its ability to shut the Strait of Hormuz and cripple global energy markets makes it difficult to defeat. There is no guarantee that the third condition can be met either, painting a dismal picture for any potential interventionist.
The stronger the moral case against a regime, the stronger the urge to just do something. But the more unilateral and externally imposed that intervention is, the heavier its burden of justification. It may still be defensible, as Kosovo shows, but the burden of justification becomes heavier because the intervention is harder to distinguish from selective power politics. Liberal politics is committed both to protecting individuals from tyrants and to allowing people to rule themselves. The answer cannot be that liberal interventionism is always wrong. The cases of Kosovo, East Timor, and Sierra Leone prove its potential, while Rwanda and Srebrenica remind us of the risks of casting it aside.
But Iraq and Libya should have ended the delusion that good intentions are all that matter. The tragedy of liberal interventionism is that the hawk sees the dissident being beaten and says: we cannot do nothing, while the dove sees the foreign army crossing the border and says: this is not your country to remake. Both are right about something. Neither has solved the problem. The tragedy is that remembering both may still not tell us what to do.