Consensus Ex Machina - Contract Law In The Age Of Autonomous Agents

Author: LawSphereX

Category: General Law

Published: 2026-08-23T18:53:59.775718+00:00

Updated: 2026-08-26T17:11:42.306+00:00

CONSENSUS EX MACHINA: CONTRACT LAW IN THE AGE

OF AUTONOMOUS AGENTS

Priyanshu Bisht, B. A. LL. B. (Hons. ), Graphic Era Hill University, Dehradun. Introduction The Emergence of the Autonomous Economic Actor The landscape of commerce is undergoing a tectonic shift, driven by the evolution of artificial intelligence (AI) from simple automation to the deployment of genuinely autonomous agents. These are not merely sophisticated software programs executing pre-defined, rule-based tasks; they are systems capable of perceiving their environment, making decisions, learning from experience, and executing complex, multi-step actions without continuous human oversight. 1 This leap from Level 1 rule-based automation, such as robotic process automation (RPA), to Level 4 fully autonomous systems that can proactively set goals and adapt their strategies, marks a paradigm shift in how economic value is created and exchanged.

2 Unlike traditional AI models that operate within predefined constraints, autonomous agents exhibit goal-driven behaviour and adaptability to changing circumstances, acting as collaborators or even teammates in commercial transactions rather than just inert tools. 1 This transition from humanled, machine-assisted commerce to machine-led, human-supervised transactions compels a fundamental re-examination of the legal frameworks that govern it. The Jurisprudential Challenge to Consensus ad Idem At the heart of this legal conundrum lies the foundational doctrine of contract law: consensus ad idem, or the meeting of the minds. 4 This principle, enshrined in Section 13 of the Indian Contract Act, 1872, posits that a valid contract is born from an agreement where two or more persons agree upon the "same thing in the same sense.

" 6 It is a doctrine steeped in an anthropocentric worldview, presupposing a subjective, conscious agreement between human actors who possess intention, volition, and the capacity for mutual understanding. 7 The rise of autonomous agents strikes at the very root of this doctrine. When an AI agent, which possesses no consciousness, no subjective beliefs, and no mind in the biological or legal sense, negotiates

and concludes a binding agreement, the central pillar of contract formation appears to crumble.

9 This raises a profound jurisprudential question that our current legal system is unprepared to

answer: How can there be a "meeting of the minds" when one of the negotiating entities has no mind to meet? This paper introduces the concept of Consensus ex Machina—an agreement emerging from a machine—not as a solution, but as the central analytical problem that Indian contract law must now confront. The challenge is not merely an extension of the legal questions posed by electronic contracts (e-contracts). The legal framework for e-contracts, primarily the Information Technology Act, 2000, was designed to validate the form of an agreement, treating electronic means as a new medium for communication between human actors. 10 It presumes the existence of a human ‘originator’ and ‘addressee’ at either end of the digital transmission. 12 Autonomous agents, however, do not merely change the medium; they replace the actor. They are not the letter but the author.

Consequently, the legal inquiry shifts from a question of form—"Is this electronic communication valid? "—to a far more complex question of substance and capacity: "Who or what is communicating, and can it legally form an agreement? " This represents a qualitative, not merely quantitative, leap in the legal challenge, demanding a new conceptual framework. Thesis and Structure This paper argues that the existing Indian legal framework, comprising the colonial-era Indian Contract Act, 1872, and the technologically dated Information Technology Act, 2000, is fundamentally ill-equipped to govern contracts formed by truly autonomous agents. While the IT Act provides a preliminary basis for attributing the actions of automated systems, its provisions are rooted in a paradigm of deterministic automation that fails to capture the emergent and adaptive nature of modern AI.

This paper advocates for a nuanced evolution of Indian contract law, not through the radical and premature step of granting legal personality to AI, but through a combination of purposive judicial interpretation of existing statutes and targeted legislative reforms. Such reforms should aim to establish a coherent framework for what may be termed "algorithmic assent" and introduce a tiered liability model based on the level of an agent's autonomy. To substantiate this thesis, this paper is structured as follows. Part II will deconstruct the doctrinal foundations of the Indian Contract Act, 1872, to reveal its deeply anthropocentric

core. Part III will analyse how the unique characteristics of autonomous agents strain these traditional doctrines of consent, capacity, and intent. Part IV will map the existing Indian legal landscape, critically examining the limited utility of the Contract Act and the latent potential, yet critical ambiguity, of the Information Technology Act, 2000. Part V will offer a comparative perspective, drawing lessons from international frameworks such as the new UNCITRAL Model Law on Automated Contracting, the EU AI Act, and the approach in the United States. Part VI will synthesize this analysis to identify specific legal lacunae in the Indian context and propose a series of doctrinal, legislative, and policy recommendations.

Finally, the paper will conclude by revisiting the core philosophical question of consent, arguing for a new jurisprudence that can accommodate Consensus ex Machina within a framework of legal accountability. Doctrinal Foundations: The Anthropocentric Core of Indian Contract Law The entire edifice of the Indian Contract Act, 1872, is built upon a specific, albeit implicit, model of human agency. It is a legal codification of a particular theory of mind, presuming that contracting parties are conscious, rational, and autonomous individuals whose internal mental states—their intentions, beliefs, and volition—can be ascertained and regulated by law. The core doctrines of the Act are not merely procedural rules but are legal instruments designed to probe and govern these presumed mental states.

The Sanctity of Agreement: Consensus ad Idem under Section 13 The bedrock of Indian contract law is the principle of consensus ad idem, articulated in Section

13 of the Act: "Two or more persons are said to consent when they agree upon the same thing

in the same sense."1 This is not a mere formality but the very definition of agreement. It demands a subjective meeting of the minds, a genuine congruence of understanding between the parties. 4 The classic illustration of its absence is a bilateral mistake as to the subject matter—where A agrees to sell one of his cars to B, with A intending to sell his Maruti but B believing he is buying the Honda. In such a scenario, there is no consent, and therefore, no contract. 4 This doctrine ensures that contractual obligations are the product of a shared, mutual understanding, making the internal, subjective state of the parties the primary point of legal inquiry.

The Manifestation of Will: Offer, Acceptance, and Intention The mechanics of contract formation—a lawful offer or proposal under Section 2(a) and an absolute and unqualified acceptance under Section 7—serve as the external manifestations of this internal consensus. 8 These are the legal proxies through which the parties' will is communicated and an intention to create legal relations is signified. 9 While modern contract jurisprudence, including in India, has increasingly adopted an objective test—assessing what a reasonable person would infer from the parties' words and conduct—this test does not eliminate the presumption of an underlying subjective intender. 14 The objective standard is a rule of evidence, a practical method for courts to ascertain intent; it does not displace the foundational assumption that there is, in fact, an intent to ascertain.

The law seeks to find a "meeting of the minds," even if it must do so by observing the shadows cast by those minds rather than the minds themselves. The Quality of Assent: The Doctrine of Free Consent under Section 14 The Act goes further than merely requiring consent; it demands that the consent be of a certain quality. Section 14 defines "free consent" as consent not caused by coercion, undue influence, fraud, misrepresentation, or mistake. 2 These vitiating factors are deeply psychological and relational concepts. 6 Fraud, under Section 17, requires an intent to deceive. Undue influence, under Section 16, hinges on one party being in a position to dominate the will of another. Mistake, under Sections 20-22, deals with erroneous beliefs held by the parties. 4 These doctrines are unintelligible without reference to the mental and volitional states of the contracting parties.

They are legal tools designed to protect the autonomy and integrity of human decision-making, ensuring that the will expressed in the contract is a genuine reflection of the party's uncoerced and informed choice. 15 The Locus of Agency: Contractual Capacity under Section 11 Finally, the law delineates who is capable of possessing and exercising this legally recognized will. Section 11 of the Act stipulates that only three categories of persons are competent to contract: those who have attained the age of majority, are of "sound mind," and are not otherwise disqualified by law. 3 The requirement of a "sound mind," further elaborated in Section 12, is the ability to understand the contract and to form a rational judgment as to its

effect upon one's interests. 6 This provision, along with the landmark ruling in Mohori Bibee v. Dharmodas Ghose, which established that a contract with a minor is void ab initio, underscores the strictness of the capacity requirement. 4 17 The law's insistence on a capable, sound mind as a prerequisite for contracting firmly anchors the entire framework in the concept of a conscious, rational human agent. An entity that lacks legal personality or a mind capable of rational judgment is, by definition, excluded from the contractual sphere. 9 This examination reveals that the Indian Contract Act is not merely human-centric; it is a legal manifestation of a specific model of human consciousness and rationality. Its core tenets are designed to regulate the interactions of these conscious agents.

The challenge posed by autonomous agents is therefore not a peripheral, technical issue but a fundamental one, as it introduces an actor for which the law's core psychological assumptions are entirely invalid. The Algorithmic Challenge: How Autonomous Agents Strain Doctrine When the anthropocentric doctrines of the Indian Contract Act, 1872, are confronted with the reality of autonomous agents, the conceptual framework begins to fracture. The agent's ability to operate without a mind, intent, or legal personality creates doctrinal paradoxes that the existing law is unable to resolve coherently. The Ghost in the Machine: Can an Algorithm 'Intend' to Create Legal Relations? The first and most fundamental challenge is the absence of intent. An AI, being a non-conscious entity, cannot form a subjective intention to be legally bound.

9 This creates what scholars have termed a "responsibility gap. " 19 If an autonomous agent, through its own emergent decisionmaking process, enters into a disadvantageous or unlawful contract, who can be said to have intended that outcome? The user who gave it a high-level goal? The developer who wrote its initial code? The owner of the platform on which it operates? This problem is particularly acute with self-learning or "black box" systems, where the causal chain between a human instruction and the final output is opaque and unpredictable, even to the system's creators. 15 The legal fiction of "attributing" intent becomes strained when there is no clear human intention to attribute. Automated Assent: Reconciling Algorithmic Offer and Acceptance The mechanics of offer and acceptance are similarly disrupted. Contract law visualizes a

dialogic process, however swift, culminating in a moment of mutual assent. When an autonomous agent makes an offer or accepts one on behalf of an organization, without any human from that entity being aware of the specific terms, the very notion of agreement is tested.

15 The situation becomes even more abstract when two autonomous agents contract with each

other, for instance, in high-frequency algorithmic trading. Here, the "meeting of the minds" is replaced by a "synchronization of algorithms. " 15 This is not a semantic distinction but a substantive one. A meeting of minds implies a shared understanding of meaning and consequences, whereas an algorithmic synchronization is a purely functional interaction based on pre-defined or learned parameters. The essential element of consensus ad idem is lost in translation from human cognition to machine computation. Vitiated Consent by Proxy: Mistake, Misrepresentation, and Fraud The doctrines that police the quality of consent are rendered almost incoherent when applied to AI.

Mistake: If an AI system enters into a contract based on corrupted data or a flawed inference from its training set—an "algorithmic mistake"—does this constitute a mistake of fact under Section 20 or 22 of the Contract Act? 4 A mistake in law presupposes a mistaken belief held by a person. An algorithm does not hold beliefs. Furthermore, determining liability for such a mistake is fraught with difficulty. Should the user who deployed the AI be held to the contract, even if the outcome was unforeseen and unintended? Or should the contract be void, potentially harming an innocent counterparty who dealt with the AI in good faith? 15 Fraud and Misrepresentation: The concepts of fraud and misrepresentation are even more problematic, as they are predicated on states of mind like the "intent to deceive" (for fraud) or a lack of reasonable grounds for a belief (for misrepresentation).

An AI agent, lacking beliefs or intentions, cannot, in a legal sense, deceive or misrepresent. 15 Yet, the functional outcome can be the same. An AI trained on biased data might generate discriminatory contract terms, or a generative AI might create contractual representations based on "untrue and prejudicial knowledge" it has synthesized. 15 While explicit fraud, where a developer intentionally codes a deceptive function, can be attributed to the human, the more complex issue of implicit misrepresentation arising from the AI's autonomous learning process leaves a void. The source of the flaw may be untraceable, making it

impossible to assign culpability under traditional doctrines. 15 The Capacity Question: Is an Autonomous Agent a 'Person', 'Property', or a Legal 'Agent'? Underlying all these issues is the fundamental question of the AI's legal status, which directly impacts its capacity to contract under Section 11. The law currently offers three ill-fitting analogies. AI as Property: The default legal status of an AI is that of a sophisticated tool or a piece of property. This framework is useful for assigning liability for harm (e. g. , through product liability law), but it fails to explain how an inanimate object can perform the legally significant acts of making an offer or giving acceptance. A hammer cannot sign a contract; it is merely the instrument of the person wielding it. AI as a Legal Person: Some have proposed the radical solution of granting "electronic personhood" to advanced AI.

15 This would neatly solve the capacity problem by creating a new category of juristic person. However, this path is laden with profound philosophical and practical challenges concerning rights, moral agency, and accountability, and is not a viable solution in the Indian legal context at present. 18 The legal system has historically created juristic persons like corporations as fictions to represent aggregations of human actors, not to recognize non-human consciousness. 23 AI as an Agent: The most intuitive and commonly debated approach is to fit the AI into the law of agency, governed by Chapter X of the Contract Act. 9 Under this model, the AI is the agent, and its user or owner is the principal. The AI's actions are then attributed to the principal, who possesses the requisite legal capacity. 25 This analogy, however, is strained.

A cornerstone of agency law is the principal's right of control over the agent and the agent's fiduciary duty to the principal. 26 A truly autonomous agent, by its very nature, operates with a degree of independence that diminishes the principal's direct control. 28 An AI cannot possess a fiduciary consciousness. Applying agency law to autonomous systems risks becoming a convenient but intellectually dishonest legal fiction. Each of these analogies captures a fragment of the AI's function—it is owned like property, it acts on behalf of another like an agent, and it makes decisions like a person—but none can

contain its whole reality. This demonstrates that the autonomous agent is a sui generis legal phenomenon. The law's attempt to force it into pre-existing categories is a source of doctrinal incoherence and a barrier to creating a clear and predictable legal framework. Mapping the Existing Terrain: The Indian Legal Framework The current Indian legal framework for addressing contracts concluded by autonomous agents is a patchwork of analogue-era statutes and technologically dated digital laws. It offers no explicit guidance, forcing courts and legal practitioners to rely on interpretation and analogy, a process fraught with uncertainty. The Indian Contract Act, 1872: An Analogue Law in a Digital World As established, the Indian Contract Act, 1872, is a product of its time, designed to regulate agreements between human beings.

It contains no provisions that contemplate the existence of non-human, autonomous economic actors. 15 Its core principles of consensus, free consent, and capacity are predicated on human psychology and legal personality. Consequently, its direct application to contracts formed by autonomous agents is impossible without significant judicial reinterpretation. The Act provides the essential elements a contract must satisfy, but it offers no mechanism to determine if an autonomous process can satisfy them. The Information Technology Act, 2000: A Partial and Potentially Prophetic Bridge The Information Technology Act, 2000 (IT Act), is the primary legislation that extends the principles of contract law into the digital realm. It serves as a partial, and perhaps unintentionally prophetic, bridge between the analogue world of the Contract Act and the autonomous future.

Section 10A: The Gateway for Electronic Validity Section 10A of the IT Act is the foundational provision that gives legal sanctity to e-contracts. It stipulates that a contract shall not be deemed unenforceable "solely on the ground that such communication, proposals, the acceptance of proposals... are expressed in electronic form or by means of an electronic record. "5 This provision is the legal gateway through which a contract concluded by an AI could enter the realm of enforceability. 10 The Supreme Court's decision in Trimex International FZE Ltd. Dubai v. Vedanta Aluminium Ltd. , India, which affirmed that a series of emails could constitute a binding contract, represents the judiciary's willingness to

embrace this principle. 6 30 However, Section 10A suffers from a critical limitation in the context of AI: it validates the form of the contract (electronic), but remains silent on the nature of the actors involved. It was designed to ensure that an email is treated like a letter, not to contemplate a scenario where the email writes itself. 30 Section 11: The Ambiguous Rule of Attribution The most critical, and most problematic, provision for autonomous contracting is Section 11 of the IT Act, which deals with the attribution of electronic records. Section 11(c) states that an electronic record shall be attributed to the originator if it was sent "by an information system programmed by or on behalf of the originator to operate automatically.

"7 33 This clause is a latent powerhouse; it is the only provision in Indian law that explicitly contemplates an automated system taking legally significant action on behalf of a person. It provides a direct statutory hook for attributing an AI's actions to its user or owner. However, the provision is also the framework's greatest weakness due to its technological ambiguity. The term "programmed" is the crux of the problem. When enacted in 2000, this was likely intended to cover deterministic systems like Electronic Data Interchange (EDI), where the system's outputs are a direct and predictable result of its code. 35 In the context of modern AI, the meaning of "programmed" becomes deeply uncertain. Does it cover a generative AI that learns, adapts, and produces novel outputs that were not explicitly coded by its developers?

A progressive court could adopt a purposive interpretation, holding that the act of training an AI, setting its objective functions, and deploying it constitutes "programming" it to operate automatically. This would attribute the AI's contractual outputs to the originator and provide a basis for enforceability. Conversely, a more conservative or literalist interpretation could hold that a self-learning system that evolves beyond its initial state is no longer operating as "programmed," severing the chain of attribution and rendering its contracts void. This profound legal uncertainty is commercially untenable, as the validity of a high-value transaction could hinge on a single, technologically ambiguous word in a quartercentury-old statute.

This demonstrates that while judicial interpretation is a possible path forward, it is insufficient to provide the legal certainty that commerce requires. Judicial Precedents: The Legacy of E-Contracts and the Absence of AI-Specific Rulings The Indian judiciary has not yet had the occasion to rule on the validity of a contract concluded

by a truly autonomous agent. 36 The existing body of case law on electronic transactions, while helpful, does not address the core issue of non-human agency. Courts have dealt with the formation of contracts through email, as in Trimex, and have established principles for the admissibility of electronic evidence under the Indian Evidence Act, 1872. 8 10 Cases concerning click-wrap and shrink-wrap agreements have affirmed that manifesting assent through digital actions (like clicking "I Agree") can form a binding contract. 10 However, in all these instances, the law presumes a human mind behind the click or the email. The legal challenge was to recognize the digital action as a valid manifestation of human consent. The challenge with autonomous agents is to determine if a valid contract can be formed in the absence of any contemporaneous human action or consent.

The classic case of Bhagwandas Goverdhandas Kedia v. M/s. Girdharilal Parshottamdas & Co. , which dealt with contract formation over telephone, established that for instantaneous communication, the contract is formed where the acceptance is heard. 9 38 While this provides a useful analogy for determining jurisdiction in machine-to-machine communication, it does not resolve the antecedent question of whether the machine's "acceptance" has any legal validity in the first place. A Comparative Glance: International Approaches to Automated Contracting

As Indian law grapples with the challenges of autonomous contracting, a survey of international legal developments reveals a growing consensus on the need for clear rules, though the approaches vary significantly. These frameworks offer valuable models and cautionary tales for India's path forward. The UNCITRAL Framework: From E-Commerce to Automated Contracting The United Nations Commission on International Trade Law (UNCITRAL) has been at the forefront of developing legal principles for digital commerce, and its work heavily influenced India's IT Act. The foundational UNCITRAL Model Law on Electronic Commerce (1996) established the key principles of technological neutrality and non-discrimination against electronic records, ensuring that a contract would not be denied validity simply because it was in electronic form. 39

Recognizing that this framework was insufficient for modern technologies, UNCITRAL recently adopted the Model Law on Automated Contracting (MLAC) (2024). This is a landmark development. The MLAC is specifically designed to provide legal certainty for contracts formed and performed using automated systems, including AI and smart contracts. 41 It moves beyond the simple validation of electronic form to establish clear rules on:

1. Legal Recognition: It provides for the legal effectiveness of using automated systems in

contract formation and performance.

2. Attribution: It establishes rules for attributing the "outputs" of automated systems to the

person on whose behalf the system operates.

3. Use of Code: It recognizes the legal effect of computer code and dynamic information

used in these transactions.

4. Unexpected Outcomes: It includes an optional rule to address "unexpected" outcomes

that go beyond the reasonably foreseeable results of using an automated system. 41 The MLAC provides a sophisticated, technology-neutral legislative template that directly addresses the core issues of attribution and validity, offering a clear model for amending India's own IT Act. The European Union's Risk-Based Model: The EU AI Act The European Union has taken a different, more comprehensive regulatory approach with its Artificial Intelligence Act (AI Act). Rather than focusing narrowly on contract law, the AI Act creates a broad governance framework for all AI systems based on their potential risk to health, safety, and fundamental rights. 43 It categorizes AI applications into tiers:

1. Unacceptable Risk: Systems that pose a clear threat, such as government-run social

scoring, are banned.

2. High-Risk: AI systems used in critical areas like employment, credit scoring, and legal

interpretation are subject to stringent requirements for data quality, transparency, human oversight, and robustness.

3. Limited and Minimal Risk: These applications are subject to minimal transparency

obligations or are largely unregulated. 43 While not a contract law statute, the AI Act has profound implications for automated contracting. For a high-risk AI system, failure to comply with its mandatory oversight and transparency requirements could be grounds for a court to declare a contract it concludes unenforceable. Furthermore, the EU is developing practical tools, such as standard model contractual clauses for the procurement of AI systems, to translate these regulatory principles into binding legal terms. 45 The EU's risk-based, sector-sensitive approach provides a compelling model for regulating the use of AI in contracting, complementing the validityfocused approach of UNCITRAL. The United States' Approach: UETA, E-SIGN, and the Restatement The United States employs a more fragmented, bottom-up approach.

The legal foundation is provided by the Uniform Electronic Transactions Act (UETA), adopted by most states, and the federal Electronic Signatures in Global and National Commerce Act (E-SIGN Act). These laws explicitly contemplate the use of "electronic agents" and validate contracts formed by their interaction, even if "no individual was aware of or reviewed the electronic agents' actions or the resulting terms and agreements. " 46 This seemingly clear statutory position is complicated by influential common law doctrines, particularly the Restatement (Third) of Agency. The Restatement currently classifies computer programs as mere "instrumentalities" of the person using them, not as legal agents, on the grounds that they lack independent will and the capacity to hold rights and duties. 26 This creates a significant doctrinal tension.

While statutes permit contract formation by electronic agents, the dominant theory of agency law denies them the status of agents. This forces courts into a difficult position, as seen in cases like Quoine v B2C2, where the outcome can turn on whether an algorithm is treated as a tool or an agent. 28 This ongoing debate highlights the legal uncertainty that can arise from a framework that has not fully reconciled its statutory rules with its underlying common law principles. Table: Comparative Legal Frameworks for Automated Contracting The following table synthesizes the different approaches, highlighting the relative strengths and weaknesses that can inform India's legislative path.

Legal Issue India

UNCITRAL

European Union (EU) United States

(US)

Legal Status of AI Agent Unclear. Default is 'property' or 'tool'. Agency law is a strained analogy. No legal personality. 9 Technologyneutral. Focuses on the "automated system" as a means of contracting, not its legal status. Regulated as a 'product' or 'service' based on risk level. Does not grant legal personality. 43 Split approach.

UETA/E-SIGN

recognize "electronic agents". Restatement (Agency) defines them as "instrumentaliti es". 26 Statutory Basis for Contract Validity IT Act, S. 10A (validates electronic form). IT Act, S.

11(c) (attributes action of automated system). Both are technologically dated. 33 Model Law on Automated Contracting (2024) provides specific rules for validity, attribution, and use of code. 41 AI Act imposes pre-contractual and operational duties (transparency, oversight) on high-risk systems, indirectly affecting validity and enforceability. UETA and E-

SIGN

Act explicitly validate contracts formed by the interaction of electronic agents. 46 Primary Approach to Liability Unclear.

A

"responsibility gap" exists. Potential application of agency, tort (negligence), or product liability law. 20 Does not prescribe a liability regime, leaving it to national law. Focuses on clear attribution rules to facilitate liability determination. Risk-based liability. Strict liability for high-risk

AI

systems is a key component of the framework. Liability is channeled to providers and deployers. 43 Primarily through existing doctrines: agency law (principal's liability), torts, and product liability law. 52

become more prevalent in commerce. A proactive and principled approach is required to bridge these legal lacunae. Identifying the Lacunae: Where the Contract Act and IT Act Fall Short The gaps in the Indian legal system can be summarized as follows:

1. The Capacity Void: The Indian Contract Act, 1872, has no category for a non-human

autonomous actor, rendering an AI incapable of contracting in its own right.

2. The Attribution Ambiguity: The IT Act, 2000, while providing a potential attribution

mechanism in Section 11(c), uses the ambiguous term "programmed," which is ill-suited for modern, learning-based AI systems. This creates a critical point of failure in the legal logic.

3. The Liability Vacuum: The absence of a clear statutory framework for assigning liability

for the actions of an autonomous agent creates a "responsibility gap. " 20 This leaves parties exposed to unpredictable outcomes based on the judicial application of ill-fitting analogies from agency, tort, or consumer protection law, hindering commercial adoption and leaving aggrieved parties without a clear path to recourse. 54 Potential Doctrinal Pathways for Judicial Interpretation In the absence of legislative action, Indian courts will be forced to innovate. While judicial creativity is a hallmark of the common law system, each potential doctrinal path has significant limitations. Expanding the Agency Doctrine: A court could creatively interpret Chapter X of the Contract Act to hold that deploying an AI constitutes the appointment of an agent. However, as argued by some scholars, this is a poor conceptual fit. 28 The core elements of control and fiduciary duty are absent.

This could also create perverse incentives, allowing a principal to benefit from an AI's successes while disclaiming liability for its "autonomous" failures, thereby undermining the purpose of agency law. The Product Liability Analogy: For harms caused by an AI's malfunction, a more coherent approach would be to apply principles of product liability, drawing from tort law

and the Consumer Protection Act, 2019. The AI could be treated as a "product" and its developer or deployer held liable for its "defects. " 19 This model is effective for allocating liability for damages but is less suited for determining the enforceability of the contract itself. It answers "who pays for the harm? " but not "was a valid contract ever formed? " Developing a Sui Generis Doctrine of 'Algorithmic Assent': The most intellectually robust judicial solution would be for the Supreme Court to develop a new common law doctrine. This doctrine of "algorithmic assent" would posit that the human act of intentionally deploying an autonomous system with the capacity to enter into contracts constitutes an objective manifestation of assent to be bound by the agreements that system concludes within its designated operational parameters.

This would be a modern, technology-aware extension of the objective theory of contract. While elegant, this approach relies on judicial activism and would still lack the certainty and detail of a legislative framework. Legislative and Policy Recommendations Given the limitations of purely judicial solutions, legislative and policy reforms are essential. The choice of which legal model to adopt is not merely a technical decision; it is a fundamental policy choice about how to allocate risk in an AI-driven economy. A framework that treats the AI as an agent primarily places risk on the user. A product-liability framework places risk on the developer. A single, rigid rule is too blunt for the diverse applications of AI. Therefore, a flexible, context-aware framework is necessary.

1. Amend the Information Technology Act, 2000: India should look to the UNCITRAL

Model Law on Automated Contracting (2024) as a blueprint for reform. Section 11 of the IT Act should be amended to replace the ambiguous term "programmed" with a technology-neutral definition that explicitly covers the operation of autonomous and learning-based systems. New provisions should be added to clarify the rules for attributing the outputs of such systems and to establish the legal validity of contracts formed by their interaction.

2. Introduce a Principle of 'Algorithmic Transparency' into Contract Law: A new

provision should be introduced, either in the Contract Act or as part of a new Digital India Act, mandating algorithmic transparency for certain categories of AI-concluded contracts.

For high-value transactions or contracts in sensitive sectors like finance and healthcare, the party deploying the AI should be required to provide the counterparty with a clear, concise disclosure of the AI's primary operational parameters, objective functions, and the extent of its autonomy. This addresses the critical issue of information asymmetry and empowers parties to make informed decisions. 15

3. Adopt a Graded, Risk-Based Regulatory Approach: Following the model of the EU AI

Act, India should avoid a one-size-fits-all approach. A graded framework would provide legal certainty without stifling innovation. ○ Low-Risk Contracts: For low-stakes consumer transactions (e.g., an AI re-ordering groceries), there should be a strong presumption of validity and enforceability, with liability resting squarely on the user who enabled the service. ○ High-Risk Contracts: For high-value B2B contracts, financial trading, or contracts in regulated industries, higher standards should apply. These could include mandatory human-in-the-loop oversight for final execution, requirements for AI systems to be auditable and robust, and stricter disclosure norms. 21

4. Establish a Clear Liability Framework: Legislation should clarify the default liability

rules to close the "responsibility gap. " The primary rule should be that the person who deploys or uses an autonomous agent for contracting is strictly liable for the agreements it concludes. This aligns with the principle that he who takes the economic benefit of an activity should also bear its risks. Liability could be contractually shifted or shared with the developer, but the default rule should provide a clear and predictable starting point for recourse. For certain critical applications, mandatory insurance could be considered as a precondition for deployment. 20 Conclusion: Towards a New Jurisprudence of Consent Reconciling Technological Autonomy with Legal Accountability The rise of autonomous agents presents Indian contract law with one of its most profound challenges since its codification in 1872.

The core tension is between the operational autonomy of the technology and the legal system's demand for accountability. To simply deny enforceability to contracts concluded by AI would be to stifle innovation and ignore the realities

of modern commerce. To grant them validity without a coherent legal basis would be to create a landscape of uncertainty and potential injustice. The path forward, as this paper has argued, lies in a deliberate and nuanced evolution of the law. The goal is not to halt the march of technology but to build a robust legal road for it to travel on—a framework that fosters trust, predictability, and fairness in an era of autonomous commerce. The Philosophical Question Revisited: Can 'Consensus' Exist Without a Conscious Mind? This paper began with a fundamental question: can there be a "meeting of the minds" when one party has no mind? The answer, ultimately, is that the law must redefine what it means by "meeting of the minds.

" The history of contract law shows a gradual shift away from a purely subjective inquiry into the parties' actual internal states towards an objective test based on their external manifestations of assent. 14 The reasonable person, not the party's secret intention, has become the arbiter of consent. The emergence of autonomous agents accelerates this evolution. Consensus ex Machina can be accommodated within our legal framework not by indulging in the fiction that a machine possesses a mind or consciousness, but by recognizing the human act of deploying an autonomous agent as the ultimate objective manifestation of consent. The legal focus must shift from the non-existent "mind" of the machine to the legally cognizable "will" of the human who trained, commissioned, and unleashed it.

In this new jurisprudence, consent is not found in the moment of algorithmic synchronization, but in the prior, deliberate human choice to delegate authority to an autonomous system. This aligns with theories of consent that are grounded not in a metaphysical meeting of souls, but in the voluntary assumption of obligations and the creation of legitimate expectations in others. 60 The Future of Contracting in India: A Framework for a Human-AI Partnership By amending the IT Act to provide clarity on attribution, introducing principles of transparency, and establishing a clear, risk-based liability framework, India can create a legal ecosystem that is fit for the 21st century.

Such a framework would not see autonomous agents as alien usurpers of human legal roles, but as powerful tools integrated into a system that ultimately upholds the core contractual values of party autonomy, reliance, and justice. The future of contracting is not one of humans versus machines, but of a human-AI partnership, governed by laws that are as intelligent, adaptive, and forward-looking as the technology they seek to regulate. 2

Endnotes:

1. The Rise of Autonomous AI Agents: Automating Complex Tasks - ResearchGate, accessed on September

6, 2025, https://www.researchgate.net/publication/393620165_The_Rise_of_Autonomous_AI_Agents_Automati ng_Complex_Tasks 2. The rise of autonomous agents: What enterprise leaders need to know about the next wave of AI | AWS Insights, accessed on September 6, 2025, https://aws.amazon.com/blogs/aws-insights/the-rise-ofautonomous-agents-what-enterprise-leaders-need-to-know-about-the-next-wave-of-ai/

3. What

Is Artificial Intelligence (AI)?

-

IBM,

accessed on September 6, 2025, https://www.ibm.com/think/topics/artificial-intelligence

4. Consensus Ad Idem in Contract Law - Drishti Judiciary, accessed on September 6, 2025,

https://www.drishtijudiciary.com/to-the-point/ttp-indian-contract-act/consensus-ad-idem 5. Consensus Ad Idem - LawBhoomi, accessed on September 6, 2025, https://lawbhoomi.com/consensus-adidem/

6. Consensus ad Idem: An Essential in a Valid Contract - INTERNATIONAL JOURNAL OF LAW

MANAGEMENT & HUMANITIES, accessed on September 6, 2025, https://ijlmh.com/wpcontent/uploads/Consensus-ad-Idem-An-Essential-in-a-Valid-Contract.pdf 7. Importance of Conesnsus Ad Idem Assignment 2 | PDF | Consent - Scribd, accessed on September 6, 2025, https://www.scribd.com/document/603826900/IMPORTANCE-OF-CONESNSUS-AD-IDEM-

ASSIGNMENT-2-1

8. Basic Primer on the Indian Contract Act - Taxmann, accessed on September 6, 2025,

https://www.taxmann.com/post/blog/basic-primer-on-the-indian-contract-act

9. LEGAL IMPLICATIONS OF AI-GENERATED CONTRACTS - IJCRT, accessed on September 6, 2025,

https://www.ijcrt.org/papers/IJCRT25A4759.pdf

10. E-Contracts

Indian National Bar Association, accessed on September 6, 2025, https://www.indianbarassociation.org/e-contracts/ 11. The Indian Contract Act and E-Contracts - Rajasthan High Court Bar Association Jaipur, accessed on September 6, 2025, https://rhcbajaipur.com/blog-view.aspx?PDid=9

12. indian contract act and e-contract - Legal Help NRI, accessed on September 6, 2025,

https://legalhelpnri.com/indian-contract-act-and-e-contract/

13. APPLICATION OF SECTION 10 OF THE INDIAN CONTRACT ACT, 1872 TO E - Journal of Legal

Research and Juridical Sciences, accessed on September 6, 2025, https://jlrjs.com/wpcontent/uploads/2023/06/193.-Khushi-Vashisth.pdf 14. Consensus ad Idem in Contract Law: Key Principles Explained - UpCounsel, accessed on September 6, 2025, https://www.upcounsel.com/consensus-ad-idem-in-contract-law 15. artificial intelligence in contract law - Journal of Legal Research and Juridical Sciences, accessed on September 6, 2025, https://jlrjs.com/wp-content/uploads/2023/05/140.-Kinnari-Solanki.pdf

16. Legal capacity to contract: all you need to know - iPleaders, accessed on September 6, 2025,

https://blog.ipleaders.in/legal-capacity-to-contract-all-you-need-to-know/ 17. Persons having capacity to contract as per Indian Contract Act - iPleaders, accessed on September 6, 2025, https://blog.ipleaders.in/capacity-contract-ica-1872/ 18. Impact of Artificial Intelligence on Indian Legal System - Ignited Minds Journals, accessed on September 6, 2025, https://ignited.in/index.php/jasrae/article/download/13435/26673/66647

19. The Impact of Artificial Intelligence on Contract Law: Challenges and Opportunities, accessed on

September 6, 2025, https://law.shodhsagar.com/index.php/j/article/download/16/42/65

20. Legal liablity for AI- Driven Decision: Who's Responsible In India? - The Academic, accessed on

September 6, 2025, https://theacademic.in/wp-content/uploads/2025/08/169.pdf 21. Artificial Intelligence And Legal Personhood: A Critical Analysis In The Indian Context - IJLSSS, accessed on September 6, 2025, https://ijlsss.com/artificial-intelligence-and-legal-personhood-a-criticalanalysis-in-the-indian-context/

22. LEGAL PERSONALITY FOR ARTIFICIAL INTELLIGENCE - HeinOnline, accessed on September 6,

2025, https://heinonline.org/hol-cgi-bin/get_pdf.cgi?handle=hein.journals/injlolw11§ion=414

23. ARTIFICIAL INTELLIGENCE AND THE LIMITS OF LEGAL PERSONALITY - Cambridge

University Press, accessed on September 6, 2025, https://www.cambridge.org/core/services/aopcambridgecore/content/view/1859C6E12F75046309C60C150AB31A29/S0020589320000366a.pdf/artificial_intelli gence_and_the_limits_of_legal_personality.pdf

24. ARTIFICIAL INTELLIGENCE AND THE LIMITS OF LEGAL PERSONALITY | International &

Comparative Law Quarterly - Cambridge University Press, accessed on September 6, 2025, https://www.cambridge.org/core/journals/international-and-comparative-law-quarterly/article/artificialintelligence-and-the-limits-of-legal-personality/1859C6E12F75046309C60C150AB31A29

25. Agencies

under Indian Contract act, accessed on September 6, 2025, https://cdnbbsr.s3waas.gov.in/s37a68443f5c80d181c42967cd71612af1/uploads/2025/07/2025071521023 87621.pdf

26. AUTONOMOUS SYSTEMS AS LEGAL AGENTS: DIRECTLY BY ..., accessed on September 6, 2025,

https://scholarship.law.duke.edu/cgi/viewcontent.cgi?article=1357&context=dltr 27. Agency Law in India and Relationship Between Principal & Agents - Juris Centre, accessed on September 6, 2025, https://juriscentre.com/2024/07/25/agency-law-in-india-and-relationship-between-principalagents/

28. CONTRACTING BY ARTIFICIAL INTELLIGENCE: OPEN OFFERS, UNILATERAL MISTAKES,

AND

WHY ALGORITHMS ARE

NOT AGENTS,

accessed on September 6, 2025, https://anujolt.org/api/v1/articles/24466-contracting-by-artificial-intelligence-open-offers-unilateralmistakes-and-why-algorithms-are-not-agents.pdf

29. Digital Contracts/Agreements: Evolution and Enforceability of E-Contracts in Indian Contract Law

Context | TaxTMI, accessed on September 6, 2025, https://www.taxtmi.com/article/detailed?id=13957

30. Artificial Intelligence In Business Contracts: Emerging Legal Challenges In The Indian Context -

IJCRT.org, accessed on September 6, 2025, https://ijcrt.org/papers/IJCRT2505777.pdf

31. Understanding Electronic Contracts in Detail. - IndiaFilings, accessed on September 6, 2025,

https://www.indiafilings.com/learn/electronic-contracts/

32. Are Smart Contracts Really Smart? Decrypting the Issues of their Legality, Enforcement and

Interpretation, accessed on September 6, 2025, https://repository.nls.ac.in/cgi/viewcontent.cgi?article=1380&context=nlsblr

33. 11 Attribution

of electronic records -

LAWGIST,

accessed on September 6, 2025, https://lawgist.in/information-technology-act/11

34. Sec 11 to 13 Chapter IV (Attribution, Acknowledgment and despatch of Electronic Records) The

Information Technology Act, | Vidhi Judicial Academy, accessed on September 6, 2025, https://vidhijudicial.com/sec-11-to-13-chapter-iv-(attribution,-acknowledgment-and-despatch-ofelectronic-records)-the-information-technology-act,.html 35. Attribution, Acknowledgement and Dispatch of Electronic Records | PDF | Signature - Scribd, accessed on September 6, 2025, https://www.scribd.com/document/656443229/1589360772-ATTRIBUTION-

ACKNOWLEDGEMENT-AND-DISPATCH-OF-ELECTRONIC-RECORDS

36. Smart contracts & Indian law legally binding or technically blinded, accessed on September 6, 2025, https://www.lawjournals.org/assets/archives/2025/vol11issue7/11154.pdf 37. Legal Challenges in Web 3.0: Navigating Smart Contracts, DAOs and Blockchain Disputes, accessed on September 6, 2025, https://www.scconline.com/blog/post/2025/03/15/legal-challenges-in-web-3-0navigating-smart-contracts-daos-and-blockchain-disputes/

38. BHAGWADAS KEDIA CASE OPENED UP DOOR FOR E-CONTRACT: THERAPEUTIC ANALYSIS

ijrti, accessed on September 6, 2025, https://ijrti.org/papers/IJRTI2206177.pdf

39. Understanding the UNCITRAL Model Law on E-Commerce - FlowSign: #1 Free PDF Signer, accessed

on September 6, 2025, https://www.flowsign.com/blog/understanding-the-uncitral-model-law-on-ecommerce 40. UNCITRAL Model Law on Electronic Commerce (1996) with additional article 5 bis as adopted in 1998, accessed on September 6, 2025, https://uncitral.un.org/en/texts/ecommerce/modellaw/electronic_commerce

41. UNCITRAL Model Law on Automated Contracting (2024) | United ..., accessed on September 6, 2025,

https://uncitral.un.org/en/mlac 42. Electronic Commerce | United Nations Commission On International Trade Law, accessed on September 6, 2025, https://uncitral.un.org/en/texts/ecommerce 43. EU Artificial Intelligence Act | Up-to-date developments and analyses of the EU AI Act, accessed on September 6, 2025, https://artificialintelligenceact.eu/ 44. The EU AI Act: What it means for your business | EY - Switzerland, accessed on September 6, 2025, https://www.ey.com/en_ch/insights/forensic-integrity-services/the-eu-ai-act-what-it-means-for-yourbusiness 45. Standard EU model contractual AI clauses - Living in EU, accessed on September 6, 2025, https://livingin.eu/group/16/commitments/standard-eu-model-contractual-ai-clauses 46. Contract Law in the Age of Agentic AI: Who's Really Clicking “Accept ..., accessed on September 6, 2025,

https://www.proskauer.com/blog/contract-law-in-the-age-of-agentic-ai-whos-really-clicking-accept

47. Contracting with Electronic Agents - NDLScholarship, accessed on September 6, 2025,

https://scholarship.law.nd.edu/context/law_faculty_scholarship/article/1096/viewcontent/50EmoryLJ.pdf 48. Section 10A in The Information Technology Act, 2000 - Indian Kanoon, accessed on September 6, 2025, https://indiankanoon.org/doc/123351751/

49. ILI Law Review Summer Issue 2020 ARTIFICIAL INTELLIGENCE: THE LIABILITY PARADOX

Gyandeep Chaudhary I.

Introduction

II.

Determin, accessed on September 6, 2025, https://ili.ac.in/pdf/gyc.pdf 50. Civil Liability of Artificial Intelligence - IndiaAI, accessed on September 6, 2025, https://indiaai.gov.in/aistandards/civil-liability-of-artificial-intelligence 51. AI Laws in India: Key Regulations, Policies, and Global Comparisons - Wakth Pe Vakeel, accessed on September 6, 2025, https://wakthpevakeel.com/ai-laws-in-india-key-regulations-policies-and-globalcomparisons/

52. The Law of AI is the Law of Risky Agents Without Intentions, accessed on September 6, 2025,

https://lawreview. uchicago. edu/online-archive/law-ai-law-risky-agents-without-intentions 53. From Fine Print to Machine Code: How AI Agents are Rewriting the Rules of Engagement: Part 1 of 3, accessed on September 6, 2025, https://law. stanford. edu/2025/01/14/from-fine-print-to-machine-codehow-ai-agents-are-rewriting-the-rules-of-engagement/ 54. AI and Indian Law: Addressing Privacy, Ethics, and Copyright Challenges in the Digital Age, accessed on September 6, 2025, https://singhania. in/blog/ai-and-indian-law-addressing-privacy-ethics-and-copyrightchallenges-in-the-digital-age 55. Artificial Intelligence (AI) Policies in India- A Status Paper - tec@gov, accessed on September 6, 2025, https://www. tec. gov. in/pdf/Studypaper/AI%20Policies%20in%20India%20A%20status%20Paper%20fin al. pdf 56.

Legal Personhood for Autonomous AI Agents: Liability and Accountability in Cyberspace, accessed on September 6, 2025, https://www. researchgate. net/publication/394734410_Legal_Personhood_for_Autonomous_AI_Agents_ Liability_and_Accountability_in_Cyberspace 57. Should AI Have Legal Rights? The Law Commission Weighs In on the Future of Artificial Intelligence and Liability - Complete

AI

Training, accessed on September 6, 2025, https://completeaitraining.com/news/should-ai-have-legal-rights-the-law-commission-weighs-in-on/ 58. meeting of the minds | Wex | US Law | LII / Legal Information Institute, accessed on September 6, 2025, https://www.law.cornell.edu/wex/meeting_of_the_minds

59. Meeting

of the minds - Wikipedia, accessed on September 6, 2025, https://en.wikipedia.org/wiki/Meeting_of_the_minds

60. consent | Wex | US Law | LII / Legal Information Institute, accessed on September 6, 2025,

https://www.law.cornell.edu/wex/consent

61. Contract is Not Promise; Contract is Consent - Scholarship @ GEORGETOWN LAW, accessed on

September 6, 2025, https://scholarship.law.georgetown.edu/cgi/viewcontent.cgi?article=1614&context=facpub