From court defeat to victory through decentralisation: TON through a lawyer’s eyes

In 2020 Telegram lost a legal battle with the SEC and returned more than $1 billion to investors who took part in the Gram token presale. It seemed the Durov brothers’ blockchain ambitions were over.

By 2024 TON had become one of the fastest-growing ecosystems, with the eponymous token integrated into the messenger, as originally envisaged. Why a courtroom defeat turned into an unequivocal market victory is explained by Sergei Ostrovskiy, a lawyer and partner at Aurum.

Background: 2018–2020

One of finance’s most eye-catching events in 2018 was Telegram’s ICO. Across two presale rounds the messenger raised $1.7bn—so much that the firm scrapped a public token sale. Buyers formed vast pools and syndicates because the minimum ticket was $20m. Even with such a giant round, the project had an oversubscription.

On October 3 2019 Telegram said the TON network and Gram token would launch by month-end. A week later, on October 11, the SEC filed suit, alleging the illegal sale of unregistered securities.

Telegram said that for “the past 18 months” it had sought clarity from the SEC about potential claims, but received only “limited” feedback.

Days later Pavel Durov’s company told investors the launch would be pushed to late April 2020. Under the SAFT, if the protocol did not launch by end-October 2019, investors were entitled to a refund. The company persuaded them to accept the delay and not seek one.

What many thought was the end came on March 24 2020, when the Southern District of New York issued a temporary restraining order blocking the distribution of Gram tokens.

On April 30 Mr Durov said the protocol and token would not launch on schedule. On June 10 2020 the dispute with the SEC was settled: Telegram agreed to pay an $18.5m fine and return $1.22bn to investors.

At the time, Telegram seemed to have lost.

How Telegram’s ICO was structured

The Gram sale was run through two firms: Telegram Group Inc. (holding company) and TON Issuer Inc. (subsidiary), both incorporated in the British Virgin Islands. In 2018 VASP rules were not yet in force there, so the choice was sensible: no regulatory constraints or obligations arose for the project.

The two-entity structure managed operational risk more effectively. It separated the risks and liabilities tied to token launch and distribution (the issuer) from the project’s core assets and funds (the holding company). The ICO structure was ring-fenced from Telegram’s core business and corporate entities—useful in the end, as the messenger and its operating companies were not parties to the litigation.

The round used SAFTs, which grant the holder the right to receive tokens in future, after the network launch. Until the Telegram case concluded, many lawyers deemed this model safe. Its logic was:

1. The token and the sale agreement (the SAFT) were treated as two separate instruments.

2. The SAFT was treated as a security issued under the Securities Act.

3. The token itself, once issued and distributed, would be a commodity, not a security.

Roughly 40 of Telegram’s nearly 200 investors were American residents, so the “US” portion relied on the 506(c) exemption, which allows sales of unregistered securities provided all buyers are verified accredited investors. Unlike 506(b), it also permits marketing and advertising.

The temporary restraining order in SEC v Telegram became the first major precedent scrutinising the SAFT structure. The court held that the initial token sale, the distribution of Grams and subsequent resales by investors formed parts of a single scheme to place tokens publicly on the secondary market (in effect, an underwriting of securities). Under the Howey test the entire scheme was an investment contract. Because the aim was the public distribution of the token—effectively an IPO—the project could not rely on Regulation D.

After the court order

The story then took unexpected turns. After the order, the TON Community Foundation declared that TON could launch without Telegram, since community consensus would suffice.

They were right. Telegram had published all code in its repository under the GNU GPL, so anyone could use it with few restrictions.

On May 12 2020 Mr Durov announced TON’s closure and stressed that his team had no ties to third-party projects building networks on the original code.

At the same time several large communities began working in parallel on their own Telegram Open Network:

1. Free TON. One of the first—and loudest—to announce plans to launch. It included TON Labs, which had worked on the original network.

Free TON launched on May 7 2020, before Mr Durov’s statement. The release followed an intriguing model: over 170 participants, including exchanges, major validators and developers, signed a declaration endorsing a “return of power to the community”, a strategy and criteria for decentralisation, a new token, TON Crystal (ticker TON), and its distribution terms.

This became one of the first DAO Product Launches, a model that gained popularity in later years as the regulatory climate tightened.

2. TON Community Blockchain. A Chinese community that planned to launch the TON token—but oddly on Ethereum. A full launch never materialised; the subsequent history is hard to trace.

3. New TON. The stealthiest of the three. Unlike the others, this group chose to build on the original TON network, whose testnet Telegram had launched before the court order. Little is known about New TON, but insiders say two key developers from the original team were involved, along with others linked in various ways to the messenger.

Almost immediately, at New TON’s initiative, The Open Network Foundation, or TON Foundation, was set up in Switzerland. It was founded by Anatoly Makosov and Kirill Emelianenko, the very developers behind the original network. TON Foundation quickly attracted a decentralised cadre of contributors.

On June 23 2021 TON Foundation published an open letter asking that the TON GitHub repository and the ton.org domain be transferred, since “Telegram no longer needs them anyway.” Unexpectedly, the requests were granted, allowing New TON to become simply TON.

Today TON Foundation works closely with Telegram; Mr Durov gives interviews promoting the blockchain, and the token is used by the messenger as an in-app currency. In effect, two years after the injunction the token was launched and integrated into the messenger, just as the Durov brothers had planned.

To me, Telegram clearly emerged as the winner. Moreover, it is reasonable to suppose that New TON was a Telegram-controlled initiative launched after the order. I have several grounds:

• key developers of the original network stood behind New TON, and many team members had ties to Telegram;
• New TON continued the original network rather than forking it;
• Telegram transferred its GitHub and the ton.org domain to New TON rather than to Free TON, whose network had already launched;
• by transferring its domain and repository, Telegram effectively recognised New TON as the successor to Telegram Open Network;
• The Open Network became integrated into the messenger, with Mr Durov himself participating in its promotion.

Legal analysis

Injunctions are almost always directed at specific persons or groups—basic enforcement practice. Once Telegram Open Network’s source code was released under the GNU GPL, anyone could launch it; forks are common in blockchain communities.

Open source let Telegram cleanly separate the product from the messenger, reinforcing the narrative that further development was handled by third-party teams that simply took code from GitHub.

The significance of Free TON

The Free TON team set an interesting precedent. It launched the network and token via a DAO—complete with a declaration-signing ceremony and an emphasis on decentralisation. That model made potential SEC claims impracticable: suing more than 170 participants scattered around the world is a legal nightmare—indeed, a non-starter.

To nitpick, Free TON’s DAO could have been structured further. Regulators might still have characterised the collective as a partnership (or simple partnership), building a case against those members who publicly identified themselves, such as Kuna, BitScale and CEX.IO. In a partnership, each member bears joint and several liability for all obligations, allowing blame to be pinned on one or a few members. Even so, that would scarcely have stopped the network launch.

The significance of New TON

New TON’s role in Telegram’s story is hard to overstate. The decision to step aside and (presumably) continue via a third-party yet controlled team was deft precisely because the injunction did not bind outsiders. The messenger was barred from distributing Grams and, in effect, launching TON; others could do so—especially after renaming Gram to TON and Telegram Open Network to The Open Network.

As validators proliferated, the network became sufficiently decentralised, which, among other things, bolstered legal safety. Telegram, for its part, was able to deploy part of the $1.7bn raised in 2018–2020 to complete initial development relatively quickly.

After settling with the SEC, the messenger raised debt via a bond placement and repaid all ICO investors, honouring its contractual obligations and the settlement terms.

Decentralisation as a shield

A DAO is not only an organisational form for Web3 communities; it is also an effective legal defence tool. Decentralisation is especially relevant for projects that cross regulatory red lines—protocols handling users’ funds: payments, lending, margin trading, DEX and DeFi in general.

Such projects often face a binary choice: shoulder the risks and launch centrally, or decentralise. DeFi protocols simply cannot operate under current licensing regimes.

Any licence or registration entails a thicket of obligations and constraints incompatible with DeFi—from broad compliance requirements to verifying every user. That makes DeFi projects especially challenging legally.

Getting decentralisation right

The preamble to Europe’s MiCA regulation says plainly that activity involving crypto-assets conducted in a fully decentralised manner without intermediaries should fall outside the regime. Those who in any way control crypto-asset-related activity, even where partially decentralised, do fall within it.

So this is not about sham DAOs run by founders or a small clique, but about organisations that are sufficiently decentralised. Western regulators understand how the industry works, can analyse transactions and have vast resources; in court a hollow DAO narrative will collapse, leaving the team and founders exposed.

Achieving sufficient decentralisation is no trivial feat. It takes serious organisational, technical and legal effort. If launching a DAO at the outset is impossible, a hybrid model can be structured with a transition to a DAO at the earliest opportunity.

A DAO “in a vacuum” does not work either. In any scenario one should plan and structure the transition in stages, and craft careful public communications. As practice shows, regulators will even listen to podcast recordings and use founders’ words as evidence.

Finally, to protect the DAO itself, it too should be given a legal wrapper. Different wrappers suit different aims.

For example, a foundation does not wrap a DAO; it typically serves as an ecosystem element that can hold assets and conduct certain operations.

A DAO LLC merges with the on-chain DAO and can recognise all tokenholders as company members, giving them corporate protections by default. Some organisations may use several structures for fuller protection.

What lessons emerge? Everyone will draw their own, but two stand out. First, a well-crafted strategy can still beat Goliath. Second, sound decentralisation really can shield a Web3 project.