AI-Assisted Contract Drafting: Legal Opportunities, Risks, and Best Practices for Businesses in India

A Legal Correspondent The inevitable application of artificial intelligence into business circles has come full circle and permeated at the core of the practice of law-the contracting process itself. India, an economy driven by the sanctity of written contracts anchored under the Indian Contract Act, 1872, further solidified through procedural compliance under the Indian Evidence Act, 1872, is seeing the swift adoption of AI-driven contract drafting tools ranging from proprietary LLMs to contract lifecycle management (CLM) technology. Beyond the allure of speed, uniformity and cost savings, AI in contracting opens an entire new vista of risks surrounding liability, confidentiality, professional standards, and even the sanctity of a contract’s intent. This article examines the numerous opportunities, critical risks, and offers a guide to best practices for businesses in India interested in deploying AI in contracts in a legal and practical manner.

I. LEGAL OPPORTUNITIES

1. Increased speed and reduced transaction costs Speed is always valued in commercial negotiations. Being able to generate first drafts, generic or standard contractual clauses, or basic contract amendments within minutes instead of hours can deliver considerable savings to organisations involved in high-volume contracting, be it under a terms & conditions agreement, a service agreement or an NDA. This is much beyond operational improvement. It frees up the time of senior, experienced lawyers to concentrate on commercial negotiations, unique risk assessment and complex custom drafting.

2. Enhanced consistency and error elimination: The most obvious reason of contract litigation in the recent past stems from poorly worded agreements such as having different values or terms in different clauses, citing wrong references or outdated clauses. Using AI based contracting solutions appropriately ensures consistency in language, identifies all occurrences of defined terms and highlights any potential omissions of legally mandatory or standard clauses. It will be extremely useful to check whether a certain required provision in commercial contracts is in place. For instance, some of the specific provisions which are legally mandated are the ones mandated under sections 7-16 of the RERA, 2016, for real estate transactions or Section 73-78 of IBC for insolvency proceedings for companies. An AI-powered tool will easily identify omissions.

3. Scalable knowledge management The organisation-specific clauses and preferred positions are lost as senior people move from firm to firm, or from client to client. An AI-driven system would be a store HOUSE of the entire organisation’s past negotiated positions on its own terms for future and better contract negotiation by new staff members and junior resources working under their guidance and control, which would save considerable knowledge management effort.

II. LEGAL RISKS

1. Unauthorised practise of law & professional liability Section 4 of the Advocates Act, 1961 states that only a ‘qualified Indian lawyer’ practising on the bar, can ‘practice law’ in India. The business of using an AI tool for drafting the contract on behalf of the own entity would not fall within the purview of the practise of ‘law’. However, in case, on how the said tool would be used for any drafting on behalf of third entities or on whose benefit such contracting tool would be used, may lead to the legal practitioners (i.e. Owners of the third party AI tool providers) or the person using such tool on their own behalf, or the professional who is taking service of such AI tool to be interpreted to be involved in ‘practice of law’. In case of an imperfectly drafted or misleading agreement, an aggrieved party could lodge claims with the developer of such AI or the owner / operator in case of a law firm or legal entity using such AI based tool in the discharge of its own professional duties, which the advocate shall be responsible for their work (as an equivalent to a firm’s paralegal, clerk or trainee).

2. Confidentiality and data protection When we are entrusting sensitive commercial/ personal data to AI systems, some of which are based out of the country and may involve transmission of data not encrypted or stored in cloud (even in some domestic cases) leads to a very complex data security issue. Such practice may expose organisations to various risks in connection with Data breach or violation of personal data protection in light of Data Protection laws such as the Digital Personal Data Protection Act, 2023 or lawyer’s obligation to their clients (as embodied in Bar Council of India rules or Evidence Act principles). In event of a data breach, strict penal provisions with damages may attract for a business. Data sharing with AI Vendors or even within its own infrastructure where access control is weak might become problematic.

3. Contract Consent and meeting of minds Every contract, under Indian Contract Act, is based on free consent of parties, meaning true ‘meeting of minds’ or ‘consensus ad idem’. An ambiguously formulated agreement, on vague data input to AI which fails to capture the parties’ exact intent, might have an argument that ‘consent was influenced’. Such claim could even lead to the contract being voided. While Indian Contract Act has not mentioned the ‘author of the contract to be human,’ an algorithmically produced contract on vague data input might invite scrutiny of intent and ‘valid consent’.

4. Intellectual Property & liability Risks Who holds copyright over content generated by an AI? Generally speaking, works created solely by an algorithm on the information already existing and derived are not protected by the Indian Copyright Act, 1957. However, AI, by drawing on other creative works on a massive scale, could infringe copyright in the very process of its creation, or may generate clauses or concepts already in public domain and there is a strong chance that AI might, by virtue of the data it is trained on, generate clause text which infringes IP rights of already known contracts. Claims relating to IP violations may then accrue against the users of such tool.

5. AI bias and discriminatory clauses It would be naive to ignore the existing biases that an AI tool, training on massive data, would inevitably acquire from that training set, particularly if biased or proprietary databases have been used. Certain clauses can become discriminatory or may have unintended discriminatory effects. Such terms may fall Foul of public policy under section 23 of the Indian contract Act. Businesses could inadvertently rely on discriminatory clauses, and face litigation from affected parties, if caught by competent judicial forums.

BEST PRACTICES FOR INDIAN BUSINESSES:

1. Ensure Human-in-the-Loop verification: All generated output must be diligently reviewed and verified by a legal professional for accuracy, completeness, and compliance with both business objectives and the governing law in India. Law firm may introduce a Partner review system to make this clear.

2. Data Confidentiality: Businesses must carefully investigate the data handling protocols and retention policies of any third-party AI contracting tool, prioritising tools with no retention, on-premise installation and clear encryption standards. Ensure strict access controls for proprietary or sensitive data, adhering to Indian Data Protection Laws.

3. Treat AI as Augmentation, not replacement: AI serves as a powerful assistant for mechanical aspects of drafting but does not replace the critical, subjective judgment of a seasoned legal professional. Leverage AI for efficiency and consistency, but rely on lawyers for nuance, complex legal strategy, and comprehensive risk assessment.

4. Ensure Human-in-the-loop Verification and Review: A qualified legal professional must thoroughly review and verify every single contract, or even any amendment thereto, generated through an AI tool to ensure its accuracy, consistency, completeness, and alignment with business objectives and all relevant Indian laws. Law firm might consider introducing partner review mandatory for AI generated contracts.

5. Data Confidentiality and Security: We’re putting our most sensitive commercial and personal information into proprietary AI platforms, many of which are based outside of our own jurisdiction or operate in cloud settings that might lack transparency and robust encryption. What happens if sensitive commercial or personally identifiable information (PII) we share, directly or inadvertently, via prompting or inputting during the drafting process gets compromised-whether through intentional breaches, data leakage during the model’s training, or inadequate access control? Companies could face stiff penalties under the Data Protection Act 2023 (DPDP Act), legal action for professional negligence due to breach of lawyer’s duty of confidentiality to clients, reputational damage, and loss of client trust.

Any business considering implementing AI in its contracting process, especially those dealing with PII or other sensitive data, should be vigilant about scrutinising AI service provider contracts and their data security practices.

Consider on-premises solutions or private cloud deployments of AI tools to maintain greater control over your data. Where possible, work with tools that have clear data minimisation policies and no retention periods for prompt inputs or AI-generated outputs, as this reduces the overall risk profile and ensures compliance with the spirit and intent of Indian data protection regulations.