I. Introduction
Artificial intelligence and blockchain are frequently described as the defining technologies of the Fourth Industrial Revolution. Although each has independently transformed distinct aspects of the digital economy, their convergence has produced legal questions that neither technology, in isolation, was capable of generating. Artificial intelligence is no longer confined to predictive analytics, natural language processing or recommendation systems. Contemporary AI models are increasingly capable of planning complex tasks, interacting with external software environments, executing financial instructions and adapting their behaviour in response to changing circumstances. At the same time, blockchain technology has evolved from serving as the underlying infrastructure for cryptocurrencies into a programmable transactional layer supporting decentralised finance, tokenised assets, digital identity, smart contracts and autonomous organisations. Together, these technologies are creating systems capable of participating in economic activity with a degree of operational independence that was previously associated only with natural or juristic persons.
This convergence is significant because it alters the role of software within commercial relationships. Traditional digital systems have largely functioned as instruments through which human decisions are implemented. Autonomous AI agents, by contrast, increasingly possess the ability to identify commercial opportunities, negotiate contractual terms, allocate digital assets, initiate blockchain transactions and execute contractual obligations without contemporaneous human approval. The legal consequences of such conduct extend well beyond questions of technological regulation. They compel a re-examination of foundational principles governing contract formation, agency, liability, evidence and regulatory accountability.
The legal system has historically been structured around the assumption that every legally significant act may ultimately be attributed to a person recognised by law. Whether commercial transactions are undertaken by individuals, corporations, partnerships or statutory bodies, the legal framework proceeds on the basis that rights, obligations and liabilities attach to identifiable legal actors capable of intention and responsibility. Autonomous artificial intelligence challenges this assumption by introducing systems capable of making operational decisions that are neither entirely predetermined nor directly supervised at the moment of execution. Blockchain technology magnifies this challenge by enabling those decisions to produce immediate legal and economic consequences through decentralised infrastructure that operates independently of conventional institutional oversight.
The purpose of this article is not to argue that artificial intelligence should be recognised as a legal person. Such proposals remain speculative and have found little acceptance within contemporary legal systems. Rather, the present inquiry examines whether existing doctrines of contract, agency and liability are capable of accommodating autonomous software systems that increasingly participate in commercial activity through blockchain networks. The central argument advanced in this article is that while existing legal doctrines remain fundamentally applicable, their application requires a more nuanced understanding of attribution, authority and responsibility in environments where autonomous systems perform economically significant acts without direct human intervention.
II. The Emergence of Autonomous AI Agents
The evolution of artificial intelligence over the past decade reflects a transition from automation to autonomy. Earlier generations of software systems operated through deterministic programming. Their outputs were limited to instructions expressly incorporated into their architecture by human developers. Automated trading systems executed transactions when predetermined thresholds were satisfied. Inventory management software reordered products after predefined stock levels were reached. Rule-based chatbots responded only to queries anticipated during system design. Although these technologies reduced the need for continuous human involvement, they remained fundamentally predictable because every operational pathway originated from explicit programming decisions.
Recent advances in machine learning, particularly the emergence of large language models combined with reasoning architectures, retrieval mechanisms and external tool integration, have altered this paradigm. Contemporary AI systems increasingly function as autonomous agents capable of pursuing objectives through adaptive decision-making rather than deterministic execution. Instead of merely responding to instructions, these systems may decompose complex objectives into intermediate tasks, obtain information from external databases, communicate with software interfaces, evaluate alternative courses of action and revise their strategies as circumstances evolve. Their conduct is constrained by objectives established by human operators, yet the specific manner in which those objectives are achieved often emerges through computational reasoning rather than predefined programming.
The distinction between automation and autonomy is legally significant. Traditional automated systems generally execute decisions that have already been made by human actors. Autonomous agents, however, increasingly determine how those objectives should be accomplished. Consider an AI procurement system authorised to minimise cloud infrastructure costs. Rather than selecting from a fixed catalogue of purchasing decisions, the system may independently compare service providers, negotiate pricing through application programming interfaces, allocate budgets across multiple vendors and execute transactions that maximise commercial efficiency. While the overall objective originates from human instruction, the individual commercial decisions may not.
This capacity for adaptive decision-making has led scholars to describe modern AI systems as agentic rather than merely automated. Agentic AI refers to systems capable of independently planning, reasoning and acting within predefined operational boundaries to achieve specified goals. The concept does not imply legal personality or unrestricted autonomy. Instead, it reflects the increasing ability of software to make context-dependent operational choices that cannot be exhaustively predicted by developers at the time of deployment. Recent academic literature has characterised this development as the emergence of an "agent economy", in which autonomous software systems interact with one another to perform economically valuable functions on behalf of human users and organisations. Such systems are expected to undertake procurement, financial management, digital asset administration and service coordination with progressively diminishing levels of direct human supervision.
The commercial significance of this development is already becoming apparent across several industries. Financial institutions are experimenting with autonomous treasury management systems capable of reallocating digital assets according to market conditions. Supply chain platforms increasingly employ AI agents to negotiate inventory replenishment, logistics and pricing. Decentralised finance protocols are witnessing the development of autonomous trading agents capable of monitoring liquidity pools, executing arbitrage strategies and optimising yield generation. These developments illustrate that artificial intelligence is evolving from a technology that merely assists commercial decision-making into one that increasingly participates in economic activity itself.
Despite these advances, autonomous AI remains fundamentally distinct from legal actors recognised by existing jurisprudence. Artificial intelligence possesses neither legal capacity nor independent rights and obligations. It cannot own property in its own name, enter contracts as a legal person or bear civil and criminal liability independent of those who design, deploy or control it. The legal significance of autonomous AI therefore lies not in its status as an independent subject of law but in its ability to undertake acts that carry legal consequences for existing persons and entities. The challenge for legal doctrine is therefore one of attribution rather than recognition.
III. Blockchain as a Transactional Infrastructure
Artificial intelligence alone does not fundamentally disrupt commercial law. Prior to the emergence of blockchain technology, autonomous software remained dependent upon conventional institutional infrastructure to produce legally significant economic outcomes. Financial transfers required banking intermediaries. Asset transfers generally depended upon centralised registries. Contractual performance ordinarily required institutional verification before legal obligations could be executed. Human oversight therefore remained embedded within the transactional architecture irrespective of the sophistication of underlying software.
Blockchain technology substantially alters this position by introducing a decentralised transactional infrastructure capable of executing transfers of value without reliance upon trusted intermediaries. Distributed ledger technology enables participants to maintain synchronised records of ownership and transactions through consensus mechanisms rather than centralised institutions. Public blockchains permit digital assets to be transferred through cryptographic authentication, while smart contracts facilitate the automatic execution of predefined contractual obligations once specified conditions are satisfied. Collectively, these characteristics enable software systems to produce legally significant economic outcomes without continuous institutional or human intervention.
The importance of blockchain extends beyond cryptocurrencies. Digital wallets controlled through cryptographic keys permit software systems to manage tokenised assets directly. Smart contracts transform contractual provisions into executable code capable of performing obligations automatically upon the occurrence of predetermined events. Tokenisation enables a growing range of assets, including securities, commodities, intellectual property interests and real-world assets, to be represented digitally on distributed ledgers. Decentralised finance protocols further permit borrowing, lending, asset exchange and liquidity management without reliance upon conventional financial intermediaries.
Artificial intelligence acquires a fundamentally different legal significance when integrated with this transactional infrastructure. Smart contracts, despite their sophistication, remain incapable of independent judgment. They execute coded instructions precisely as written but cannot interpret commercial context, negotiate contractual terms or evaluate unforeseen factual circumstances. Autonomous AI supplies precisely these capabilities. An AI agent may analyse market conditions, negotiate pricing, determine whether contractual prerequisites have been satisfied and subsequently instruct blockchain infrastructure to execute transactions through smart contracts. Blockchain therefore provides the mechanism through which autonomous decision-making is translated into legally consequential conduct.
This relationship between artificial intelligence and blockchain is complementary rather than competitive. Artificial intelligence introduces adaptive reasoning into commercial systems, while blockchain supplies verifiable execution, immutable records and decentralised settlement. Together they create the technical conditions necessary for autonomous economic participation. Software no longer functions merely as an intermediary between human intention and commercial execution. Instead, it increasingly occupies an operational position within commercial relationships themselves, exercising delegated discretion over assets, contractual performance and financial transactions.
The legal implications of this transformation extend considerably beyond technology law. Once autonomous software acquires the practical ability to negotiate, transact and execute legally significant obligations through decentralised infrastructure, traditional legal doctrines governing consent, authority, attribution and liability inevitably require reconsideration. Existing legal principles were developed within a commercial environment in which economically significant acts could ultimately be traced to identifiable human decision-makers. The emergence of autonomous AI operating through blockchain networks requires an examination of whether these principles remain adequate when software itself becomes the immediate actor through which commercial transactions are concluded.
IV. The Rise of Machine-to-Machine Commerce
The integration of autonomous artificial intelligence with blockchain infrastructure has facilitated the emergence of a new form of commercial activity in which software systems transact directly with one another. Electronic commerce has historically remained human-centric despite technological mediation. Online marketplaces, digital payment systems and algorithmic trading platforms have undoubtedly automated significant aspects of commercial activity, yet the legally relevant decisions continued to originate from identifiable human actors. The technology executed instructions that had already been formulated by individuals or organisations. Autonomous AI agents operating on blockchain networks depart from this model by introducing the possibility that commercial interactions may occur between software systems acting within delegated authority but without contemporaneous human participation.
Machine-to-machine commerce is not merely a theoretical construct. Autonomous agents are increasingly capable of identifying counterparties, negotiating commercial terms, selecting payment mechanisms, initiating blockchain transactions and verifying contractual performance through decentralised infrastructure. In decentralised finance ecosystems, for instance, AI agents may monitor liquidity pools, compare lending rates across protocols, rebalance digital portfolios and execute transactions in response to changing market conditions. Similar developments are emerging within decentralised supply chain management, cloud computing marketplaces and tokenised asset platforms, where software agents allocate resources, purchase computing capacity and administer digital assets through blockchain-based payment systems.
The legal implications of these developments extend beyond questions of technological capability. Traditional commercial law presumes that contractual negotiations involve parties capable of intention, consent and legal accountability. Autonomous agents challenge this assumption because they increasingly undertake commercial acts through computational reasoning rather than deterministic programming. An AI procurement agent authorised to minimise infrastructure costs may negotiate subscription terms with an AI pricing agent representing a software provider. Once mutually acceptable conditions are identified, settlement may occur automatically through blockchain infrastructure and contractual obligations may immediately become enforceable through smart contracts. Although both systems ultimately operate on behalf of identifiable principals, the commercial interaction itself may occur entirely between autonomous software systems.
The emergence of decentralised autonomous organisations has further complicated this landscape. Unlike conventional corporations, DAOs rely upon blockchain-based governance mechanisms through which decisions are implemented by smart contracts following community voting or algorithmic processes. Artificial intelligence is increasingly being integrated into these governance structures to analyse proposals, optimise treasury management and execute operational decisions. The combination of decentralised governance and autonomous decision-making introduces significant challenges concerning attribution and responsibility because the decision-making process becomes distributed across multiple technological and organisational layers.
The economic significance of machine-to-machine commerce should not be underestimated. The World Economic Forum has repeatedly identified autonomous digital agents as a foundational component of future digital markets, while technical standard-setting organisations have begun examining governance models for trustworthy AI agents operating across decentralised ecosystems. Commercial practice is therefore advancing more rapidly than the legal doctrines that govern it. The law is increasingly required to address situations in which software does not merely facilitate commercial relationships but actively participates in them.
V. The Emerging Research Gap
Legal scholarship concerning artificial intelligence and blockchain has largely evolved along parallel trajectories. The literature on artificial intelligence has predominantly focused upon algorithmic discrimination, explainability, data governance, automated decision-making and regulatory oversight. Questions relating to transparency, accountability and the ethical deployment of machine learning systems have understandably occupied a central position within contemporary discourse. At the same time, scholarship concerning blockchain technology has concentrated upon cryptocurrencies, decentralised finance, digital assets, smart contracts and the legal implications of distributed ledger technology. Although both bodies of literature have developed rapidly, comparatively little attention has been devoted to the legal consequences arising from their convergence.
This separation is increasingly difficult to justify. Artificial intelligence without blockchain generally lacks an independent transactional mechanism through which commercial decisions may be implemented. Conversely, blockchain without artificial intelligence remains dependent upon predetermined programming or direct human instructions. Their convergence fundamentally alters the legal character of both technologies. Artificial intelligence becomes capable of producing legally significant commercial outcomes, while blockchain evolves from a passive transactional infrastructure into an environment within which autonomous decision-making may be exercised.
Recent academic work has begun recognising this convergence. Scholars have examined decentralised AI architectures, blockchain-based identity frameworks for autonomous agents and governance mechanisms for machine-to-machine transactions. Research has also explored the use of decentralised identifiers, verifiable credentials and cryptographic reputation systems to establish persistent identities for AI agents operating within blockchain ecosystems. These developments seek to enhance trust, accountability and interoperability while preserving the decentralised characteristics of blockchain infrastructure. Nevertheless, much of the existing literature remains technological rather than doctrinal in orientation. Questions concerning contract formation, agency, negligence and liability continue to receive comparatively limited analysis, particularly within the context of Indian law.
The position is even more pronounced in India. Although the Information Technology Act, 2000 recognises electronic records and electronic contracts, it was enacted at a time when autonomous artificial intelligence did not exist in its present form. Similarly, the Bharatiya Sakshya Adhiniyam, 2023 modernises the evidentiary treatment of electronic records but does not specifically address immutable blockchain ledgers or autonomous software systems. The Digital Personal Data Protection Act, 2023 establishes obligations concerning automated processing of personal data, yet its principal focus remains privacy and data governance rather than commercial attribution or transactional liability.
The absence of technology-specific legislation should not be interpreted as a legal vacuum. Existing principles of contract, tort, agency and corporate law continue to apply unless displaced by statute. The challenge lies in determining whether these principles can adequately accommodate autonomous systems whose operational behaviour cannot always be reduced to deterministic programming. This question becomes particularly significant where blockchain transactions are irreversible, smart contracts execute automatically and commercial losses may arise within seconds of an autonomous decision being taken.
Comparative developments suggest that regulators have begun acknowledging these concerns. The European Union's Artificial Intelligence Act adopts a risk-based approach centred upon obligations imposed on providers and deployers rather than recognising AI as a legal subject. The United Kingdom Law Commission has examined the enforceability of smart contracts and the legal treatment of digital assets while emphasising that existing common law possesses considerable flexibility to accommodate technological innovation. UNCITRAL has similarly recognised the legal validity of automated electronic transactions through its work on electronic commerce and transferable electronic records. These developments indicate an emerging international preference for adapting existing legal principles rather than creating entirely new legal categories for autonomous technologies.
VI. Towards a New Legal Inquiry
The convergence of autonomous artificial intelligence and blockchain does not require a rejection of established legal doctrine. Contract law has repeatedly adapted to new methods of communication, from handwritten correspondence to telegraphic communications, electronic contracting and automated online transactions. Agency law has accommodated increasingly sophisticated commercial relationships involving employees, independent contractors, electronic systems and algorithmic trading platforms. The history of commercial law demonstrates that legal principles frequently outlast the technologies through which commerce is conducted.
The present challenge, however, differs in one important respect. Previous technological innovations altered the manner in which legally significant acts were communicated or executed. Autonomous artificial intelligence has begun altering the manner in which those acts are formulated. Software is no longer confined to implementing decisions that have already been made by human actors. It increasingly participates in identifying commercial opportunities, evaluating alternatives and selecting courses of action that ultimately produce legal consequences. Blockchain infrastructure ensures that those decisions may immediately affect legal rights through the transfer of digital assets and the execution of smart contracts.
This distinction has significant implications for private law. If autonomous systems increasingly determine how commercial objectives are achieved, questions concerning consent, authority, attribution and responsibility acquire renewed importance. Existing doctrines remain relevant precisely because the law continues to attribute rights and obligations to natural and juristic persons. The difficulty lies not in recognising artificial intelligence as a legal actor but in determining the circumstances under which the conduct of autonomous systems should be attributed to those who design, deploy, supervise or benefit from them.
The discussion that follows therefore examines whether the existing doctrines of contract and agency are capable of addressing these emerging forms of commercial activity. Rather than asking whether autonomous artificial intelligence should be treated as a contracting party, the more useful inquiry is whether traditional principles governing offer and acceptance, delegated authority, apparent authority and attribution remain sufficiently flexible to govern commercial relationships increasingly mediated by autonomous agents operating through blockchain infrastructure. It is within this framework that the next part of this article considers contract formation, autonomous execution and the allocation of authority under existing legal doctrine.
At Fathom Legal, we advise businesses, founders and technology companies on AI governance, blockchain transactions, digital asset regulation, smart contracts and emerging technology law, helping clients innovate with legal certainty in an evolving regulatory landscape. If your organisation is building with AI or blockchain, our team can help design legally resilient frameworks from the outset.