Tokenising a Real-World Asset Does Not Create a New Regime

ADGM's guidance on digital securities is precise about the mechanism. Where a digital or virtual token being offered is assessed to exhibit the economic and legal features and characteristics of a Security, the FSRA will deem it a Security pursuant to section 58(2)(b) of the Financial Services and Markets Regulations. It is a deeming provision, not a new category, and it does its work on the substance of the claim rather than on the label the offer carries. Once that door is passed, the ordinary securities regime applies — offers, prospectus obligations, the market rulebook, the conduct rules, the prudential category — and none of it is modified by the fact that the register is a distributed ledger.

The same guidance's taxonomy is worth reading for what it excludes as much as for what it captures. Derivatives over digital assets and units in funds investing in them are regulated as ordinary specified investments. Virtual assets and utility tokens are treated as commodities and are not specified investments. Privacy tokens and algorithmic stablecoins are prohibited by statute for use in carrying on any regulated activity. Five destinations, and the classification cascade has to be run before anything else can be decided, such as whether it involves licensing inside a designated economic zone.

There is also a hard limit that surprises people. ADGM's virtual-asset framework does not cover initial token offerings or capital formation, and the guidance adds that capital formation activities are not provided for under that framework and are not envisaged under the market rulebook. A virtual-asset permission is not a route to raising money in a token. Securities-like tokens go down the securities route or they do not go, which may influence the decision to pursue the onshore route licensed by DED.

The Capital Market Authority's published virtual-asset guidelines place digital securities and DLT-securities outside the virtual-asset framework: they are considered securities subject to the legislation regulating traditional securities, plus whatever technology requirements the regulator imposes. Digital commodity derivative contracts are treated the same way, which may require understanding what is UAE trade licence.

The framework the regulator issued on 13 April 2026 then does something specific. Its Alternative Trading System module is not virtual-asset-only — the regulator's own release states that it extends to conventional multilateral trading facilities for securities and to multilateral trading facilities dedicated to tokenised securities. The stated organising principle for the whole framework is same activity, same risk, same regulatory outcome, which is the shortest available statement of why tokenisation is not its own regime, and why choosing where the licence should sit is important.

One caution, stated plainly because it affects how much weight the reader should give any of this. The operative resolution behind the April 2026 framework is not retrievable from the regulator's published pages, and the older guidelines describing a six-licence regime remain on the same site. The eight activity names are public; the definitions, perimeters and transition arrangements are not. A separate draft regulation on security tokens and commodity tokens was consulted on in January 2025 and we have found no published issuance notice for it. Anyone telling you exactly which track a tokenised instrument sits under is reading something we have not been able to find.

Deciding that an instrument is a security only starts the structuring conversation. The next question is what the entity actually is, and the published regimes contemplate several answers that are not interchangeable.

Tokenised vehicles are not hypothetical in the UAE. ADGM announced the Realize T-BILLS Fund as the first tokenised T-bill fund in ADGM, and its own framing of tokenisation is forward-looking: at Abu Dhabi Finance Week in December 2025 the FSRA presented tokenisation, decentralised finance and AI-driven market participation as the next wave the framework must respond to, rather than as an existing dedicated regime.

That is the right way to read every live example. It is an existence proof that the route works. It is not a template, because what is published is the outcome, and what decided the outcome — the drafting of the claim, the choice of vehicle, the sequencing of approvals — is not published by anyone.

Dubai's regulator has issued two public notices on this subject. In April 2025, jointly with the Dubai Land Department, it addressed entities falsely claiming or otherwise purporting to participate in the pilot phase of the DLD Real Estate Tokenisation Project, stating that the project was officially launched on 19 March 2025 as a limited pilot phase involving select participants approved by both bodies. In February 2026 it issued an update concerning recent market communications and representations relating to participation in the pilot and the offering of tokenised real estate products in or from the Emirate.

We will say nothing about that pilot beyond what those notices state, and neither should anyone else. The relevant fact for a reader considering a real-estate tokenisation structure is not what the pilot permits. It is that this is a topic where imprecise public claims have already drawn regulatory attention, which is a reason to take advice rather than a route to copy.

None of the above is portable. Dubai excluding the Dubai International Financial Centre is VARA's perimeter, by VARA's own definition of the Emirate. DIFC is the Dubai Financial Services Authority's, and is a separate regime we do not describe here — the DFSA's own site is not reachable to us, and we will not characterise a regulator's rules from a page we could not open. ADGM is the FSRA's, geographically bounded. Every other emirate, including its commercial free zones, is the federal regulator's, and financial free zones are excluded from the federal virtual-asset regime entirely.

Four regimes, four vocabularies, four sets of activity names, and a token classification that has to be run separately in each of them. Choosing the jurisdiction before choosing the regime is the wrong order, and it is expensive to unwind because the entity is fixed at the first approval.

Which of several tracks a tokenised instrument sits under — conventional securities law, the tokenised-securities venue module, a fund regime, or a virtual-asset permission — is a re-licensing-level decision. The drafting choices that decide it are made at design stage and are effectively fixed afterwards, because the claim the token carries is what is being classified and rewriting the claim after issuance is a different transaction entirely.

That is the judgement, and it is not available on any regulator's website. It is settled by reading a specific instrument against several regimes at once, deciding which entity ought to exist before any of them is approached, and accepting that some designs will need to change to be licensable at all.

The classification that decides everything here is made at drafting stage, in the terms of the claim the token carries, and it is effectively fixed once the instrument is issued. No regulator publishes the answer for a specific instrument, and one of the operative federal texts is not in public circulation at all.

So bring us the claim before the chain: what the holder is actually owed, by whom, and on what terms. We will read it against the regimes that could take it, tell you which vehicle it needs and in what order the approvals have to be sought, and be straight with you about which parts of the design would have to change to be licensable. This is structuring commentary rather than legal advice, and it is not an opinion on how any regulator will read a particular set of facts.