New Reality: Buterin Creates the Future While the State Writes Laws — Analysis of the Anonymous Bulletin Board on Aztec

New Reality: While the State Writes Laws, Buterin Is Already Creating the Future. Analysis of the Anonymous Bulletin Board Project on Aztec

Author: Editorial Board of “Kafedra” / SforNews

July 21, 2026

While regulators in Europe, Vietnam, and Russia are trying to find levers of control over cryptocurrencies, Ethereum co-founder Vitalik Buterin simply takes and creates tools before which these laws will prove powerless. His new project, which he called a “toy demo version,” is more than just a technical experiment. It is an architectural shift that changes the rules of the game.

Our colleague from Crypto Emergency, Yan Krivonosov, has already drawn attention to this [https://t.me/crypto_yan/359], and we are grateful to him for an excellent overview. We will look at this event through the lens of our editorial optics — as a practical implementation of the principles we have been discussing for several months now.

Protocol vs. Law: Buterin Shows That “You Cannot Prohibit What You Cannot Control”

In our materials, we have repeatedly emphasized: bitcoin and similar protocols were created as systems beyond the control of states. Buterin’s project is direct proof of this thesis.

What did he do? He created on the Aztec blockchain [1] — an L2 solution with a priority on privacy [8][15] — an anonymous bulletin board [2][9]. A user deposits ETH onto the first layer of Ethereum, anonymously writes a post on the second layer, and then withdraws the funds back. The sender’s address never “appears” in open data, and the connection between the deposit and the message is completely broken thanks to zero-knowledge proofs (ZK-proofs) [2][9]. The code is already openly available on GitHub [2].

This is not just an abstract “idea of anonymity.” It is a working prototype of an infrastructure that technically cannot be blocked or censored from the outside. While the State Duma is holding readings of bills on digital currencies, and Vietnam is preparing to introduce fines, Buterin is simply showing: “Your laws do not work in this environment.” This is precisely the case when technology is moving far ahead, and attempts to regulate it using old methods are doomed to fail.

The Paradox of AI Moderation: A Solution or a New Problem?

The key element of the project is a built-in AI moderator that scans text and automatically bans “immoral” advertisements [2]. At first glance, this is an elegant solution to the problem of moderation in a decentralized environment, which Yan mentioned in his post, recalling the experience of creating a decentralized social network.

Indeed, the use of a local LLM (the Qwen3.5-2B model [3][10], running through llama.cpp) for automatic filtering is a step forward compared to cumbersome systems of voting or oracles. However, a new challenge lies here as well.

Who determines what is “immoral”? AI, unlike a state censor, does not know borders. The model was trained on a specific dataset that carries the bias of its creators. Who will set the criteria for the neural network? It turns out that we are simply transferring the function of censor from a person to an algorithm, which is also not neutral. As we write in our materials, old problems are not solved — they take on a new form. And this paradox is one of the key ones for discussion in the coming era.

The AI Arms Race: When Algorithms Fight Algorithms

Buterin’s project is just the tip of the iceberg. We are entering an era where AI agents are becoming key players. And this is where the real race begins.

On the one hand, tools are emerging that can be used for protection and investigations. Chainalysis has already announced its “blockchain intelligence agents” [4][11], capable of compressing multi-day investigations into minutes, and SlowMist has launched TrackAgent [5][12], which helps victims of fraud track stolen funds across 31 blockchains.

On the other hand, the same or similar tools in the hands of attackers can be used to automate phishing and search for vulnerabilities. This is a classic arms race where the one with the faster and smarter algorithm will win.

And here our analysis fully coincides with the conclusions of Crypto Yan: the struggle is no longer so much between hackers and analysts, but between those who adapt AI to their tasks faster. And this completely changes the landscape of crypto security.

“Vibe Coding”: A New Level of Access to Creating Reality

The method by which the prototype was created deserves special attention — “vibe coding,” popularized by Andrej Karpathy [6][13]. Buterin, in essence, described the desired behavior in simple language, and AI tools generated the code [2][9].

This means that the barrier to creating complex decentralized applications is rapidly falling. You no longer need to be a super-programmer to deploy a protocol with ZK-proofs. An idea and the ability to explain it to a neural network are enough [13]. This is a tectonic shift that will lead to an explosive growth in the number of decentralized social networks, anonymous bulletin boards, and payment systems that simply cannot be blocked. A corporation or government will no longer be able to delete your account — it will live on the blockchain.

Architectural Conclusion: The Future Is Not Created in Parliaments

Yan Krivonosov’s material and this Buterin project are not just news. They are an illustration of our main thesis: the old world is dead, and the rules in the new world are written not by politicians, but by code and algorithms.

While states argue about taxes on stablecoins and licenses for miners, architects like Buterin are building infrastructure that makes these arguments meaningless. And, as we see in the example of the FMBA, where AI helps create personalized cancer vaccines in 42 days [7][14], the pace of technological change is not just high — it is exponential. And it is impossible to stop it.

Sources

[1] Aztec Network. “Announcing the Alpha Network” (March 30, 2026) — description of the launch of the first fully decentralized L2 network with privacy on Ethereum.

[2] Vitalik Buterin deploys 3 zero-knowledge prototypes on Aztec Network // KuCoin (July 19, 2026) — details of the anonymous bulletin board project, vibe coding, and three prototypes.

[3] meeTARA – Qwen3.5-2B-Base (GGUF) // Hugging Face — the local AI model used for moderation in Buterin’s project.

[4] Chainalysis Launches First AI Blockchain Intelligence Agents // Chainalysis (March 30, 2026) — announcement of AI agents for investigations and compliance.

[5] TrackAgent: The On-Chain AI Intelligence Agent // SlowMist (July 9, 2026) — launch of an agent for tracking stolen assets across 31 blockchains.

[6] Andrej Karpathy: From Vibe Coding to Agentic Engineering // BAAI (May 2, 2026) — the concept of vibe coding and its evolution into agentic engineering.

[7] Skvortsova: AI helps find mutations for creating cancer vaccines // RT Russian (June 3, 2026) — on the use of AI in the FMBA for personalized vaccines.

[8] The Aztec Stack // Aztec Network (May 25, 2026) — technical description of the Aztec architecture.

[9] Why Isn’t Vitalik Pumping ETH? His Vibe-Coded Billboard Asks Too // Yahoo (July 19, 2026) — details of the project and Buterin’s demo post.

[10] Qwen3.5-2B GGUF // Hugging Face — the Qwen3.5-2B model in GGUF format.

[11] Chainalysis to Add Blockchain Intelligence Agents // CoinMarketCap (April 1, 2026) — on the integration of agents into the Chainalysis platform.

[12] TrackAgent AI Agent Joins SlowMist’s Lost Fund Recovery // KuCoin (July 9, 2026) — technical details of TrackAgent.

[13] Tesla’s Former AI Director Andrej Karpathy on vibe coding anniversary // Times of India (February 4, 2026) — overview of the evolution of vibe coding over the year.

[14] Head of FMBA revealed the role of AI in creating cancer vaccines // NEWS.ru (June 4, 2026) — on the timelines for creating personalized vaccines.

[15] Aztec Network: Roadmap Update // Aztec Network (March 3, 2026) — roadmap update and technical details.

Link to Crypto Yan’s material: https://t.me/crypto_yan/359

© The material was prepared by the editorial board of the journals “Kafedra” and SforNewsbased on open data and partner analytics. When citing, a reference to the original source is mandatory.