A ghost flight needs an aircraft, a crew, a registration, a permissive ramp, and cargo. It also needs something less obvious and considerably harder to improvise: a legal person willing to sign.
Somebody has to hold the lease. Somebody has to be the party to the charter agreement, the name on the air waybill, the account holder at the bank, the insured under the policy, the entity that invoices the customer and receives the payment. None of that can be done by a man. It has to be done by a company, because that is what a company is for, and the entire apparatus of international commerce is built to transact with legal persons and to refuse to transact with anyone else.
Which means the binding constraint on this industry is not secrecy. It is the availability of counterparties, manufactured to order, at a price, with as little public disclosure as the transaction can tolerate.
The United Arab Emirates built the factory, and it did so for reasons that had nothing to do with any of this.
The zone as a product, not a loophole
Jebel Ali Free Zone opened in 1985 next to a new deepwater port, with nineteen companies. The proposition was straightforward and, at the time, unusual: foreign investors could own their businesses outright, pay no corporate tax, import and re-export without customs duty, and repatriate profits in full, inside a legally demarcated area with its own authority and its own rules.
It worked spectacularly. JAFZA now hosts thousands of companies including a substantial roster of global corporates, the adjacent port became one of the busiest on earth, and the model was copied across the country until the UAE had dozens of free zones, each with its own registrar, licensing regime, and fee schedule. Each zone is, in effect, a small jurisdiction with its own rules operating inside a federal state, which multiplies the number of doors an investigator has to knock on and the number of authorities who have to agree before anything happens.
It is important to state plainly that this is a legitimate and enormously successful economic development strategy. A state with hydrocarbons, a small population, and an ambition to become a trading hub identified the frictions that deter foreign business and removed them. The overwhelming majority of the entities in these zones are ordinary firms doing ordinary things.
The externality is that one of the frictions removed was disclosure, and disclosure is exactly the friction that a deniable logistics operation is willing to pay a premium to avoid.
Nobody designed the free zone to service arms trafficking, any more than Ilyushin designed a freighter for it. The zone was built to make company formation fast, cheap and confidential for legitimate business, and those three properties are equally useful to everybody else. The pattern is now familiar enough to state as a rule: in this trade, the instruments were almost always built by somebody respectable for a respectable purpose, and the grey market inherited them rather than commissioning them. A proprietary airline, a Soviet freighter, a free port and an offshore registry were each built by serious institutions for serious reasons, and each turned out to have a second use nobody costed.
The UAE free zones are a price list
The span from Jebel Ali to Ras Al Khaimah is not a route. It is a spectrum, and understanding it as a spectrum is the single most useful thing in the subject.
At the Dubai end sits JAFZA and the other established Dubai zones. Expensive, institutionally credible, attached to real port and warehouse infrastructure, accepted by major banks, and capable of holding Dubai freehold property through the Land Department. A counterparty examining a JAFZA entity sees a recognised jurisdiction with substance behind it.
At the northern end sits Ras Al Khaimah. RAK International Corporate Centre was formed by a 2016 merger of two earlier registries and is now the largest offshore registry in the country by number of entities, with over twenty thousand registered. Formation runs from roughly seven to nine thousand dirhams a year, which is on the order of two thousand dollars. It requires a registered agent rather than an office. Multiple share classes are permitted, there is no minimum capital, shares transfer easily, and a company can elect to operate under English common law rather than local law. Beneficial ownership details are not published.
Further north still, Ajman offers something cheaper again, with the notable limitation that banks in most cases decline to open accounts for its companies.
Read that as a curve and the trade is obvious. Moving from Dubai toward Ras Al Khaimah, cost falls, disclosure falls, formation speed rises, and institutional acceptance falls with them. Moving the other way buys credibility with a counterparty at the price of visibility and money.
An operator picks the point on that curve that matches the least trusting party it has to satisfy. If the bank is the obstacle, pay for Dubai. If the bank has already been solved some other way and the only requirement is a name on a contract, buy the cheapest entity that will hold a signature. Most operations use several at once, at different points on the curve, matched to the counterparty each one has to face. The bank sees the respectable entity. The charter contract for a flight into a contested airspace is signed by something that costs less to lose.
What the entity actually accomplishes
The function is worth separating from the mystique, because the mystique is misleading.
A free zone company does not hide anything by existing. It performs four specific jobs.
It creates a contracting party. An aircraft lease, a charter, a ground handling agreement and a fuel contract all require a legal person on each side, and a newly formed entity is a legal person from the day the licence issues, with no operating history required and no track record that anybody can consult.
It creates a payment channel. An invoice from a licensed company to a customer is an ordinary commercial transaction, and a company with a bank account can receive it. Where banking proves difficult, the settlement moves outside the banking system entirely, which is a separate problem with its own infrastructure.
It creates separation. The entity signing the charter need not be the entity that owns the aircraft, which need not be the entity that employs the crew, which need not be the entity that took the customer’s money. Each of those can be a different company in a different zone, and establishing the relationship between them requires a formal process in each.
And it creates deniability of a particular legal shape. The beneficial owner of a company is not a party to the company’s contracts. If the entity is named in a report, the entity is what has been named.
That last property is the whole business. Everything else in this industry is a variation on the same manoeuvre performed with a different instrument: a flag relocates the aircraft’s legal identity, a registry relocates the regulator, and a corporate vehicle relocates the responsible person.
Companies are disposable and aircraft are not
Here is the asymmetry that shapes everything downstream, and it is a matter of arithmetic.
An Il-76 in flyable condition costs millions of dollars, takes time to acquire, requires maintenance and a type-rated crew, and cannot be replaced quickly. A free zone company costs two thousand dollars, forms in days, and can be abandoned without consequence.
So the two behave completely differently under pressure. Name an operator in a United Nations panel report and the company can be dissolved and reconstituted under a new name within a month, with the same aircraft, the same crews and the same customers. Name an aircraft and the aircraft still exists, with the same serial number, wherever it has been repainted to.
Which is why anybody serious about tracing this trade works from tail numbers rather than corporate names, and why corporate designations have such limited effect. Designating a company removes a label. The capability underneath is untouched and relabels itself at the cost of a formation fee. That asymmetry between the durability of assets and the disposability of identities is not unique to aviation, and it is unusually stark here because the aircraft are so scarce and the companies so abundant.
The panels of experts that name these entities understand this perfectly well, which is why the good reports list registrations alongside company names.
The register exists. It is just not public
The most common misconception about this subject is that beneficial ownership in the UAE is unknown. It is not unknown. It is unpublished, and those are entirely different things.
Cabinet Resolution 58 of 2020, as subsequently amended, requires all UAE companies including offshore entities to maintain an up-to-date register of ultimate beneficial owners, defined as any natural person owning twenty-five percent or more of shares or voting rights, or otherwise exercising control. That register is filed with the relevant authority, whether RAK ICC, JAFZA or another zone.
It is not publicly accessible.
The distinction matters enormously for how this industry actually operates. The information exists in a filing cabinet. A UAE regulator can see it. A foreign regulator can request it through mutual legal assistance, which takes months and requires a predicate. A journalist, a researcher, an NGO or a targeted-sanctions analyst cannot see it at all.
So the product being sold is not concealment from the state. It is concealment from everybody else, which is cheaper to provide, easier to defend as a legitimate privacy interest, and sufficient for almost every purpose an operator has. The parties who would act on the information are precisely the parties who cannot obtain it, and the party that can obtain it has limited incentive to act.
That structure also explains why the reforms of recent years have improved compliance metrics without changing much operationally. Requiring a register and not publishing it satisfies the standard while preserving the product.
The FATF episode
The clearest test of all this came from the international standard-setter, and the outcome is instructive in both directions.
The Financial Action Task Force placed the UAE on its list of Jurisdictions under Increased Monitoring, the grey list, on 4 March 2022, citing strategic deficiencies in countering money laundering and terrorist financing. The underlying 2020 mutual evaluation had identified limited supervision of non-financial businesses including precious metals dealers, inadequate risk assessment, weaknesses in international cooperation, and insufficient implementation of targeted financial sanctions.
The response was fast and substantial. The UAE created a dedicated executive office, revised legislation, expanded money laundering investigations and prosecutions, extended supervision to designated non-financial businesses, improved beneficial ownership transparency mechanisms, and strengthened information exchange.
On 23 February 2024 the UAE was removed from the grey list, in under two years, which is less than half the average time taken to achieve delisting.
Then the objection, filed the same day. An investigative organisation noted that the removal came a month after a leaked United Nations Panel of Experts report cited credible evidence that the UAE was supplying arms to warring parties in Sudan, in the face of an arms embargo dating to 2005, and argued that a framework substantially improved on paper had not obviously changed conduct in practice.
Both things are true simultaneously and the tension between them is the point. The compliance architecture genuinely improved. The specific conduct continued. A standards body assesses systems, because systems are what can be assessed, and a system can be excellent while the thing it was built to prevent carries on beside it.
That is the same finding the British government produced in 1976 about an airline it had correctly identified and could not stop, restated in the language of financial regulation.
Why the model persists
Three structural reasons, none of which involves anybody behaving badly.
The first is competition. There are dozens of free zones in the country and more across the region, each with revenue targets and a registrar whose performance is measured in registrations. A zone that imposes materially more disclosure than its neighbours loses business to its neighbours. That is a race whose direction is set by the incentive structure rather than by anybody’s intentions, and it is the same dynamic that governs registries of every kind.
The second is that the legitimate demand is overwhelming and genuine. Confidentiality is wanted by family offices, by firms with competitors, by individuals in jurisdictions with kidnapping risk, and by ordinary businesses that simply do not want their ownership structure published. A policy that eliminated confidentiality to catch the small fraction misusing it would destroy an enormous legitimate business, and no government does that voluntarily.
The third is that the model is the economy. The UAE’s position as a trade, logistics and financial hub rests substantially on being the easiest place in the region to establish and operate a company. That is not a side effect of the development strategy. It is the development strategy, and it succeeded on its own terms to a degree that makes any reversal politically unthinkable. The economic logic that produced it is sound, the results are visible from orbit, and the externality is a rounding error against the gains as measured by the state that collected them.
Where the free zone sits in the stack
Set the components beside each other and the completeness of the arrangement becomes visible.
The aircraft carries a registration from one state. The operating company is licensed in a free zone in a second. The owning company may sit offshore in a third. The crew hold licences from a fourth and passports from a fifth. The flight departs a ramp in a sixth, and the cargo originates in a seventh.
Not one of those arrangements is unlawful on its own. Each is a routine commercial choice with a conventional justification, and each is made in a jurisdiction with no obligation to consider the others. A regulator anywhere in that chain examining its own fragment finds nothing wrong, because there is nothing wrong with the fragment.
Assembling a complete picture therefore requires cooperation across six or seven legal systems simultaneously, several of which earn revenue from not cooperating, none of which has jurisdiction over more than its own fragment, and all of which can honestly state that the element within their jurisdiction complies with their law. That is the coordination problem the whole subject keeps arriving at, expressed as a corporate structure rather than as a flight plan.
That is the architecture. It was not designed by anybody. It emerged because each state optimised its own regulatory offering independently, and the gaps between the optimisations turned out to be a usable space.
Substance, and the test nobody applies
There is a concept in tax and regulatory law called substance, and its treatment here explains why the reforms of recent years changed less than they appear to.
Substance asks whether an entity does anything. Does it have employees, premises, decision-making capacity, and activity commensurate with the income it books, or is it a name on a certificate? The UAE introduced economic substance requirements in 2019 under international pressure, applying to entities conducting specified activities including holding company business, shipping, and distribution.
The requirements are real and they are calibrated. A pure holding company faces a reduced test. An entity that books no relevant income in a period has limited obligations. And the assessment is made by the authority that licensed the entity and collects its fees.
The practical consequence for an operator is that substance is purchasable at the level required. A desk, a resident director, a filed return and a modest local spend satisfy a great deal, and the cost of doing so is small relative to a single charter.
What substance testing is genuinely good at is catching the entirely artificial arrangement with no local footprint at all. What it is not designed to do, and does not do, is establish what an entity with adequate substance is actually shipping. A company with an office, a director and audited accounts can charter an aircraft to an airfield in eastern Chad and remain fully compliant with every substance requirement in the jurisdiction, because the requirements concern the company’s reality rather than its cargo. Substance tests whether the entity exists. Nothing in the framework tests what it does.
The claims that do not hold up
An audit, because this subject attracts lazy characterisation from both directions.
UAE free zones are money laundering vehicles describes a tiny fraction of a system overwhelmingly used for legitimate commerce, and treating the whole as the exception is both unfair and analytically useless.
The UAE does not know who owns these companies is false. Beneficial ownership registers are required and filed. They are not public.
The FATF delisting proves the problem is solved conflates a framework assessment with an outcome, which is what a framework assessment is not.
The FATF delisting was political ignores substantial documented reforms and a technical process with published criteria.
Shell companies are inherently illegitimate misdescribes a holding company, which is an ordinary and lawful instrument used by essentially every large enterprise in the world.
Designating companies stops the trade is contradicted by the arithmetic. A designated company costs two thousand dollars to replace.
The free zones were created for this purpose inverts the chronology. JAFZA opened in 1985 for reasons entirely unconnected to any of this, and the use came later because the properties were useful. The same chronological error gets made about flags of convenience, which predate their notorious applications by decades.
Publishing beneficial ownership would end the problem overstates it, since the aircraft can already be tracked and the enforcement gap is jurisdictional rather than informational. Publishing would raise costs and change who bears them, which is worth something and is not the same as ending anything.
What the free zones are actually telling us
The finding worth carrying is that this industry’s hardest component to replace is not the aircraft, the crews or the cargo. It is the willingness of some jurisdiction to produce a legal person on demand and decline to publish who stands behind it.
Aircraft are finite and ageing. Crews are a shrinking pool of men trained by a state that no longer exists. Cargo depends on conflicts that come and go. But corporate identity is manufactured, unlimited, and costs about as much as a business-class ticket, and the supply expands to meet demand because forming companies is a fee-generating business and the fee is collected whoever pays it. The scarce inputs in this industry are physical and finite. The abundant one is legal and manufactured, and the abundant one is the one that defeats enforcement.
Which inverts the intuitive hierarchy. The romantic version of this subject is about pilots and aircraft in dangerous places, and the pilots and aircraft are the visible part. The load-bearing component is a registered agent in an office park north of Dubai filing an incorporation document, for a fee, on behalf of a client he has met once, recording the beneficial owner in a register he is obliged to keep and not obliged to show to anybody who might ask. The gap between what is recorded and what is accessible is the entire product.
The investigation ends at an airport in Darfur, and the paper trail leading to it runs through a filing cabinet in a jurisdiction that is not obliged to open it.
A company is cheaper than a landing fee. That single fact explains more about how this industry survives enforcement than any amount of analysis of the aircraft, because you cannot designate your way out of a supply that regenerates for two thousand dollars a time.