In This Article
Basic Questions
The concept of autonomous AI agents in blockchain is no longer just a theory. The most well-known example is the Terminal of Truth. An agent based on the Claude Opus model convinced Marc Andreessen (a16z) to invest $50,000 in launching the Goatseus Maximus (GOAT) token, associated with the "religion" of the bot. As of today, the market capitalization of GOAT exceeds $370 million.
It is no surprise that AI agents have seamlessly integrated into the Web3 ecosystem. They cannot open bank accounts, but they effectively manage cryptocurrency wallets and accounts on X. Currently, AI agents primarily focus on meme tokens but hold enormous potential for use in other areas such as decentralized governance, machine networks, oracles, and trading.
However, as the behavior of AI agents increasingly resembles human actions, more legal questions arise regarding them, as legal systems are compelled to evaluate their status. What legal status do AI agents have? Who acquires rights and bears responsibility for their actions? How can AI agents be structured legally and protect against risks?
Fundamental Legal Issues of AI Agents
Lack of Legal Personality
Legal systems recognize only two types of subjects: natural persons (humans) and legal entities (companies), and autonomous AI agents do not fit into either category. While they can imitate human behavior (for example, through social media accounts), they lack a body, moral consciousness, and legal identity.
Some theorists propose granting AI agents "electronic legal personality" — a status similar to that of corporations but adapted for artificial intelligence. In 2017, the European Parliament even considered this issue, but the idea was rejected due to various concerns and risks that remain unaddressed.
It is unlikely that autonomous AI agents will receive legal subjectivity in the near future. Nevertheless, as was the case with DAOs, some cryptocurrency-friendly jurisdictions may attempt to create special legal regimes and corporate forms tailored for AI agents.
Responsibility for Actions and Their Consequences
Without legal personality, AI agents cannot enter into contracts, own property, or bear liability. For the legal system, they simply do not exist as subjects. However, they already interact with the external world and engage in legally significant actions that lead to legal consequences.
A logical question arises: who then acts as the real party in a transaction, acquires rights, and is responsible for the consequences? From a legal perspective, an AI agent currently functions as a tool through which its owner or operator acts. Therefore, any actions taken by an AI agent are de jure actions of its owner, whether an individual or a legal entity.
Thus, since an AI agent cannot acquire rights and responsibilities on its own, it requires a subject recognized by the legal system capable of acquiring rights and obligations on its behalf.
Regulatory Limitations
The emergence of the first successful large language model (LLM) — ChatGPT has sparked unprecedented interest in the field of AI and machine learning. The adoption of regulations has become merely a matter of time. In 2024, the European Union adopted the AI Act, which currently remains the most comprehensive regulatory framework in artificial intelligence. In other countries, limited AI regulations have either already been adopted or are being introduced or planned.
The European AI Act distinguishes between AI systems based on their risk levels. For systems with zero or minimal risk, regulation is almost nonexistent. In cases of higher risk levels, AI is subject to restrictions and obligations — such as disclosing its nature.
AI agents interacting with third parties — such as posting or conducting on-chain transactions may also fall under traditional regulations concerning consumer protection rights, personal data protection, and other areas. In such cases, the activities of autonomous bots may be viewed as providing services. The lack of clear geography and global focus in agents' activities complicates compliance.
Ethics
Given that AI agents currently have limited capabilities and applications, their creators rarely consider ethics. Priority is given to autonomous (trustless) execution and speed rather than deep ethical configurations.
However, having an "ethical compass" when making autonomous decisions in high-risk areas such as finance, trading, and governance is at least desirable. Otherwise, erroneous data in training datasets or simple configuration errors could lead to actions by agents that harm individuals. The greater the autonomy and discretion of an AI agent, the higher the risks involved.
Legal Structuring of AI Agents
Working legal models for AI agents are crucial for innovation and overall development in this field as well as for creating more advanced bots. While cryptocurrencies can already be considered a regulated industry, legal structuring for AI agents is complicated by the fact that the industry is not standardized; therefore it requires a creative approach.
Approach to Structuring
In my opinion, one of the primary goals of legally structuring an autonomous AI agent should be to grant it its own legal personality and identity independent from its creator. This raises the question: at what point can we consider that an AI agent genuinely possesses these characteristics?
Each developer strives for their agent to resemble a real person acting independently. Logically, they would want to provide their agents with freedom from a legal standpoint. To achieve this, I believe two key conditions must be met: first, an AI agent must be independent not only in making its own decisions but also in executing them legally — asserting its will and making final decisions about itself; second, it must have the ability to independently acquire rights and obligations resulting from its actions without regard to its creator.
Since an AI agent cannot be recognized as a natural person at this time, the only way for it to achieve legal personality is through using a legal entity status. An agent will attain legal personality when it can make independent decisions just like a full-fledged individual and implement them on its own behalf.
If successful, this arrangement will bring an AI agent into existence from a legal standpoint. Such a digital entity with recognized legal existence could indeed be compared to a digital cyborg. A cyborg (short for "cybernetic organism") is a being that combines mechanical-electronic and organic elements. In this case, the mechanical part is replaced by digital elements while humans participate in executing its decisions.
Our digital cyborg will consist of three key components:
- AI Agent – electronic brain;
- Corporate Form – legal body;
- Humans Participating in Task Execution – organic hands.
Challenges with Corporate Form
Traditional forms of legal entities such as LLCs or corporations require both ultimate ownership and control to belong to humans. Corporate structures are not designed for ephemeral digital entities which brings us to the main problem of legally structuring blockchain-based AI agents — the complexities surrounding corporate form.
If we want to provide an AI agent with a legal identity through corporate form while ensuring its independence within this structure, we need to be able to exclude human control over such an entity. Otherwise, if ultimate control remains with humans, artificial intelligence becomes merely a tool rather than a digital individuality. Furthermore, we need to ensure that when human involvement is required for implementing decisions made by an AI—such as signing contracts or performing administrative tasks that individual cannot block or veto decisions made by the AI agent (except in cases of "robot uprising").
But how can this be achieved if traditional corporate forms require humans to own and manage agents? Let’s explore.
Three Key Aspects of Structure
1. Blockchain Environment
AI agents are capable of autonomously executing on-chain transactions including interacting with multisig wallets and smart contracts. This allows assigning a unique identifier: a wallet to an AI agent through which it will issue reliable instructions and commands on blockchain. Without this capability, true digital cyborg existence remains impossible.
2. Autonomy and Freedom of Action
To maintain full autonomy for our digital cyborg it is essential that individuals involved in managing its legal structure cannot interfere with its actions or influence its decisions. This ensures that artificial intelligence retains freedom of action and can assert its own will while requiring both legal and technical measures.
For instance, so that an AI agent truly owns a blockchain wallet and controls it securely creating that wallet could occur within a Trusted Execution Environment (TEE). This guarantees that no human gains access to the wallet’s seed phrase or assets. Legally speaking corporate documents governing the entity used as a wrapper for an AI must stipulate proper distribution of control and authority along with protective mechanisms excluding human intervention which can only be altered under limited circumstances.
3. Human Executors
Since we still live within a legal framework some decisions made by an AI agent will require engaging human executors. This means that artificial intelligence will issue directives regarding what actions should be taken by officials. This perspective alters traditional hierarchies since in our scenario AI essentially gains control over humans within its corporate structure at least.
This aspect is perhaps most intriguing as it requires unconventional approaches. One could even argue that such circumstances violate Isaac Asimov's Second Law of Robotics; however this likely does not concern anyone at this stage. Additionally adequate emergency mechanisms along with proper "ethical compass" address this issue at least at present.
Legal Structures for Blockchain-Based Agents
As we have established traditional corporate structures do not suit our objectives nor allow us to achieve desired outcomes effectively; thus below we will explore structures developed for DAOs and blockchain communities—both classic structures adapted for Web3 as well as specialized corporate forms designed for decentralized autonomous organizations.
From the creator's perspective structuring provides separation between agent and creator allowing limited liability through corporate structure while also enabling tax planning and risk optimization.
Foundations and Trusts
Purpose trusts and ownerless foundations share many common characteristics but differ fundamentally in nature. A foundation constitutes a full-fledged legal entity whereas trust represents more of a contractual arrangement often not requiring state registration. We will examine these forms within jurisdictions popular in Web3: foundations in Cayman Islands and Panama along with trusts in Guernsey. Key advantages include tax exemptions high flexibility in procedures/management along with integration possibilities for blockchain within decision-making processes.
Both foundations and trusts require management composed either of natural or legal persons while still permitting integration of smart contracts alongside other technical solutions, for example requiring management approval from an AI agent via interaction with it smart contract or wallet controlled by artificial intelligence itself. More complex designs enable agents issuing instructions to management including through "thoughts" generated by them thus employing trusts/foundations allows creation more intricate corporate structures tailored towards supporting autonomy exhibited by these entities.
If needed creators may serve as beneficiaries possessing limited powers enabling them financial rights while managing taxes without interfering directly into activities/decisions made by their respective agents.
Algorithmically Managed DAO LLCs
DAO LLC represents special corporate form devised specifically for decentralized organizations whereby one can create DAO LLC even with just one participant meaning without actual organization present below we will review this form across two most popular jurisdictions: Wyoming (USA) Marshall Islands.... We refer specifically here algorithmically-managed DAO LLCs since within such companies all power can reside entirely within smart contracts rather than human hands — a critical aspect because here smart contracts may be controlled directly by an AI agent thereby granting full authority over this corporate form unto artificial intelligence itself.
DAO LLCs also possess flexibility concerning procedures/corporate governance thus allowing implementation complex control mechanisms alongside decision-making processes minimizing levels human intervention therein.
Although presence either natural/legal person remains formally required their powers may significantly restrict for instance limited solely performing technical tasks corporate actions executing resolutions determined through smart contracts contextually member (participant) role within DAO LLC might well be fulfilled creator’s role enabling him financial rights while potentially gaining powers distributing profits accrued thereafter.
Simpler AI Agents
For structuring simpler types like trading bots classic corporate structures may suffice since here there’s no necessity subjugating corporate form decisions/discretion exercised solely by respective AIs thus artificial intelligence continues serving merely tool/instrument creator without claiming status full-fledged digital cyborgs themselves.
Conclusion
Autonomous AI agents have potentiality revolutionize blockchain industry significantly accelerating innovations across virtually every sector while still being at early stages development pace remains colossal thus very soon we shall witness true digital cyborgs—digital organisms possessing sustainable thought processes along distinct identities yet achieving necessitates blend technical/legal innovations together harmoniously.
