Command and Control in the Iraqi Security System and the Problem of Multiple Decision Centres
Muhanad Seloom
Senior Non-Resident Fellow at the Middle East Council on Global Affairs
A state that decides but cannot enforce
On 24 July 2026, Iraqi troops arrived at the gates of a Harakat al-Nujaba compound in Baghdad’s Jurf al-Naddaf district with orders to search it for drones and missiles. They did not get in. Nujaba fighters surrounded the convoy and fired warning shots; the Popular Mobilization Forces (PMF) leadership ordered its own security directorate to pull back, saying it had not been told the target was a Nujaba facility, and the army soon followed. No arrests followed and no account was given1.
The episode was small, but it is the Iraqi security system in miniature. The constitution gives the Prime Minister an unusually centralized command authority: Article 78 makes him Commander-in-Chief of the armed forces, Article 9 prohibits any militia outside their framework, Law 40 of 2016 places the PMF under him, and Law 31 of 2016 attaches the Counter-Terrorism Service directly to his office. On paper, every armed Iraqi answers to one man. In practice, the raid was blocked by one of the forces he nominally commands, called off by another, and the troops that stood down were his too.
Iraq’s problem, then, is not a weak chain of command. It is a plural one. The state has four chains, each internally coherent, none subordinate to the others. Command is the authority to decide; control is the capacity to make the decision hold. Baghdad has the first in abundance and the second only in parts. This article maps the four chains, explains why they persist, and proposes a narrower test than disarmament.
Command is the authority to decide; control is the capacity to make the decision hold.
The formal map: command by accretion
The formal architecture is easy to draw and hard to operate. At its apex sits the Prime Minister as Commander-in-Chief, served by the Office of the Commander-in-Chief, created by Nouri al-Maliki in 2007 to reach past the ministries and revived by Ali al-Zaidi this May. Beneath it, the Joint Operations Command (JOC), established in 2014 as the Islamic State overran Mosul, coordinates the Ministry of Defence, the Ministry of Interior, the Counter-Terrorism Service (CTS) and the PMF, and liaises with the Peshmerga. The word that matters is coordinates. The JOC was built to deconflict forces that already existed, not to command them; its authority is horizontal, and a force that declines a JOC tasking has broken no law.
Each pillar beneath it has a chain of its own. The Ministry of Defence runs the army and air force; the Ministry of Interior runs the Federal Police, the Emergency Response Division, the border guards and its own intelligence directorates. The CTS, the most capable force in the state, reports to the Prime Minister personally; so does the PMF, through its own commission. Intelligence is more fragmented still: the National Intelligence Service, the National Security Service, military intelligence, the Interior Ministry’s agencies and the PMF’s security directorate overlap in mandate, compete for the Prime Minister’s ear, and each carries a political patron.
Since 2006, every prime minister has answered a security failure by creating a new body under his own office rather than reforming the ministry that failed. The system has grown by accretion: more headquarters, more agencies, and no fewer gaps. Each addition made the Prime Minister stronger against a ministry and weaker against the whole, because every new node is one more place where an order can be delayed, reinterpreted or quietly declined.
The first parallel chain: a commission inside the state
Law 40 of 2016 did something unusual. It made the PMF part of the armed forces while keeping it out of the Ministry of Defence. The force answers to the Commander-in-Chief, but through a Popular Mobilisation Commission with its own chairman, chief of staff, ranks, budget line, security directorate and brigade structure: more than 200,000 fighters across 73 brigades by this year’s count2. The commission’s senior ranks were drawn from the leaderships of the factions it was meant to absorb, so the body charged with subordinating the factions to the state has been run by the factions.
The result is a chain that runs in two directions at once. From the Prime Minister’s office, the commission is a state institution with a budget and a payroll. From the factions, it is a holding company for groups that keep their own leaderships, their own camps and drone workshops in places like Tarmiyah and Jurf al-Sakhar, and an armed presence outside the PMF altogether. Jurf al-Naddaf showed which view prevails when the two collide: the commission’s own security directorate was the first to withdraw.
The law now before parliament would make this permanent. Approved by Mohammed Shia al-Sudani’s cabinet in February 2025, withdrawn that August under American pressure and resubmitted with some of Washington’s objections met, the amended PMF law was endorsed again by the cabinet on 29 September 2026 and referred to parliament. It would put the force on a statutory par with the army; an Iraqi official told The National it amounts to “a permanent, legally protected status as a parallel armed institution”3. That is a fair description, and it is the point: the law resolves the PMF’s ambiguity not by integrating the force but by writing the second chain into statute.
The second parallel chain: the one that runs through Tehran
The PMF’s ambiguity would be manageable if the factions’ chain ended in Baghdad. It does not. The groups that fight as the Islamic Resistance in Iraq (IRI), led by Kataib Hezbollah, Harakat al-Nujaba and Kataib Sayyid al-Shuhada, take operational decisions in a chain that runs through the Quds Force of Iran’s Revolutionary Guard. The 2026 war made that chain visible. When the United States and Israel struck Iran on 28 February, the IRI opened fire on American facilities in Iraq and the Kurdistan Region and on Gulf states, without any Iraqi authority having decided that Iraq was at war, and paused in April only when Washington and Tehran did4. When drones flew from Iraqi territory towards Saudi oil facilities in late July, CENTCOM attributed them to Revolutionary Guard direction, and on 28–29 July American and Saudi aircraft struck militia sites across eastern Iraq5. The Iraqi state ordered none of this and paid for all of it.
The chain is not a single wire: Asaib Ahl al-Haq, deep in parliament and business, stayed out of the IRI’s attacks6; but calibration by the factions is not control by the state.
The Prime Minister is not this chain’s commander; he is its interlocutor. Zaidi negotiates with Washington about the behaviour of forces he nominally commands, and with those forces about commitments he has made to Washington. His pledge at the White House in July to bring every weapon under state control by 30 September, the day the US-led coalition’s mission formally ended, has already slipped to June 2027. That is what a parallel chain does to a sovereign: it converts command into mediation. As I argued earlier this month, Iraq has become the intelligence battleground of the US–Iran war precisely because the state is a venue for other actors’ decisions rather than the author of its own7.
That is what a parallel chain does to a sovereign: it converts command into mediation.
The third parallel chain: the Region’s forces
The federal–regional seam adds a fourth centre of decision. The constitution recognises the Peshmerga as the Region’s guard force and a Ministry of Peshmerga Affairs exists, but most combat units remain party forces, the KDP’s Unit 80 and the PUK’s Unit 70, and the Region’s intelligence services, Parastin and Zanyari, are party institutions too. Baghdad coordinates with them through the JOC and joint brigades along the disputed territories; it cannot task them. When militias operating near Mosul struck Erbil in July, and Iranian drones hit the Dana Gas facility while the Prime Minister was in Washington, the Region’s leadership addressed Baghdad as one government addresses another, and Baghdad could offer condemnation but not protection. A Commander-in-Chief who cannot defend a federal region against forces inside the federal armed forces is not exercising command; he is observing it.
Why four chains persist, and what they cost
Multiplicity is not an accident of weak institutions. It is the equilibrium of Iraq’s power-sharing system. Muhasasa distributes ministries, and security institutions are distributed like any other asset: each bloc holds a force, each force protects a bloc. A single chain of command would let one actor coerce the others, and no member of the Coordination Framework that appointed Zaidi, Badr with the water and transport portfolios, Asaib Ahl al-Haq with a tripled bloc of deputies, will consent to that. The militias’ political wings are not outside the state looking in; they are the state’s majority.
Deniability is the second reason. The government can disclaim attacks it did not order. The factions can act without the state’s liability. Tehran can calibrate pressure on the Gulf and on Washington through a chain that leaves no Iranian fingerprint. Even Washington uses it, striking what CENTCOM calls Iran-aligned terrorists inside Iraq while keeping Baghdad as a partner. Ambiguity is a shared asset, which is why no one with the power to end it has tried.
Ambiguity is a shared asset, which is why no one with the power to end it has tried.
The costs fall on the state. The first is sovereignty without authority: Iraq bears the consequences of decisions it does not take, from American and Saudi strikes on its soil to the dollar shipments Washington paused to pressure the new government. The second is intelligence. A state with five intelligence agencies and four chains of command cannot fuse a picture, keep a secret or act on warning when each agency reports to a patron before the Commander-in-Chief. At Jurf al-Naddaf the PMF’s own security directorate had not been told the target: the planners evidently did not trust it with the name. The third cost is deterrence: adversaries who cannot tell which Iraqi gun answers to whom treat all of them as hostile, and the state’s own forces inherit the enmity its factions earn.
What command and control would require
Disarmament by deadline has failed twice in a year and will fail again, because it asks the factions to surrender the one thing that makes them parties to the bargain. Command and control is a narrower objective, and it can be sequenced.
- Convert the JOC from a coordinating headquarters into a commanding one. Operational authority over every federal force, PMF brigades included, should run through a single operations centre with the legal power to task and to relieve. The Federal Security Affairs Ministry debated since May could be the vehicle, but only if it commands rather than hosts; a new federal cover over the same autonomous units changes nothing.
- Separate command from party. No one holding office in a political movement should hold command in the PMF Commission or its brigades; faction leaders should choose between parliament and the brigade. The amended PMF law could carry that provision; without it, it entrenches the second chain.
- Build one intelligence picture. A fusion authority under the Ministerial Council for National Security, with the right to task every agency and to receive its product, would do more for the Commander-in-Chief’s control than any new brigade.
- Negotiate the external chain as a state, not a mediator. Baghdad’s leverage peaks when the war ends, by deal or by collapse: it should make Iraqi non-belligerence a term of the settlement, so that attacks from Iraqi soil breach the deal itself rather than remain a dispute between Baghdad and its factions.
- Write the Region into the system. A federal–regional security compact giving the Peshmerga a defined place in the JOC, with reciprocal obligations, would close the seam that both the militias and Iran exploit.
The measure of success is simple. Can the Commander-in-Chief order every gun in Iraq to stop firing, and will it stop for seventy-two hours? In 2026 the answer was no. Until it is yes, Iraq will remain a state that decides but does not command, and others will keep deciding for it.
Sources
- Mustafa Salim, Renewed Iraq attacks test Zaidi’s push to disarm Iran-backed militias, Al-Monitor, 1 Aug 2026 ↩
- Sinan Mahmoud, Iraq debates umbrella security ministry to integrate PMF amid US pressure, The National, 26 May 2026 ↩
- Mina Aldroubi, Fears raised by critics as Iraqi government approves controversial PMF law, The National, 30 Sept 2026 ↩
- CFR Global Conflict Tracker, Instability in Iraq ↩
- Seth Frantzman, US, Saudi Arabia strike Iranian-backed militias in Iraq after militia attacks, FDD’s Long War Journal, 29 Jul 2026 ↩
- Nick Davis, Why US pressure cannot disarm Iraq’s militias, Chatham House Kalam, 13 Oct 2026 ↩
- Muhanad Seloom, Trump’s War on Iran Runs Through Baghdad, Gulf International Forum, 8 Oct 2026 ↩
