January 29, 2025: The Madras High Court Enforces Strict Enablement Limits on AI Inventions
On January 29, 2025, the Madras High Court issued a landmark decision in Caleb Suresh Motupalli vs Controller Of Patents that fundamentally reshaped the patent landscape for artificial intelligence applications in India. The court ruled that AI-related patent claims fail the enablement test under Section 10(4)(a) if they lack specific technological implementation and require undue experimentation. This decision establishes a legal precedent that abstract AI concepts, without detailed structural features and operational data, will face immediate rejection under the Patents Act, 1970.
The ruling came at a critical juncture as the Indian Patent Office simultaneously released its 2025 Guidelines for Examination of Computer Related Inventions (CRIs), codifying the court's reasoning into formal examination protocols. Both the judicial decision and administrative guidelines now demand that applicants provide practical, reproducible details rather than theoretical extrapolations or conceptual frameworks.
Section 10(4)(a) Enablement Test: What AI Patent Applicants Must Prove
The enablement test under Section 10(4)(a) of the Patents Act, 1970, requires that a complete specification must describe the invention in a manner sufficiently clear and complete to enable a person skilled in the art to work it without undue experimentation. In the context of AI and machine learning inventions, this statutory requirement translates into a demanding disclosure standard that goes far beyond conceptual descriptions. The court determined that a software engineer with expertise in AI and allied fields constitutes the Person Skilled in the Art (PSITA) for evaluating AI/AR inventions.
The Madras High Court found that the applicant's persona-extender invention failed precisely because the complete specification lacked technological implementation details and required excessive inventive effort to reproduce. The court stated: "For the aforementioned lack of technical criteria in the complete specification to work the claims for achieving the intended result, the claimed invention fails the enablement test under Section 10(4)(a) of the Patents Act."
Section 10(4)(b) and the Requirement to Disclose the Best Method of Performing the Invention
Beyond basic enablement, Section 10(4)(b) requires disclosing the best method of performing the invention known to the applicant at the filing date. This provision creates an additional layer of risk for AI patent applicants who may be tempted to file early-stage applications before fully developing operational prototypes. The 2025 CRI Guidelines emphasize that disclosure must reflect practical implementation rather than theoretical possibilities, meaning applicants cannot sustain an application based on aspirational or projected capabilities.
The distinction between theory and practice proved fatal in the Caleb Suresh Motupalli case, where the specification described AI functionalities without providing the structural features, training datasets, algorithmic parameters, or validation metrics necessary for a skilled practitioner to implement the system. This gap between claimed results and disclosed methods triggered both Section 10(4)(a) and Section 10(4)(b) objections.
Section 3(k) Exclusions: When Abstract AI Concepts Remain Unpatentable
Section 3(k) of the Patents Act, 1970, explicitly excludes mathematical or business methods, computer programs per se, and algorithms from patentability. The 2025 Guidelines clarify that abstract AI concepts and mathematical models remain patent-excluded unless integrated into tangible physical or technical applications with explicit operational data. Mere conceptualization of AI capabilities, without demonstrating a workable 'what' and 'how' through structural features, results in immediate rejection.
The legal framework now requires applicants to bridge the gap between abstract computation and concrete technical effect. In practice, it is observed that successful AI patent applications must identify specific hardware configurations, data preprocessing steps, network architectures, or sensor integrations that transform a general-purpose computing device into a specialized technical apparatus.
The Judicial Standard Borrowed from C Van Der Lely NV v. Ruston's Engineering Co. Ltd.
The Madras High Court drew upon the principles established in C Van Der Lely NV v. Ruston's Engineering Co. Ltd., a 1993 decision from the United Kingdom, to interpret the enablement requirement. That case emphasized that a specification must enable a person skilled in the art to perform the invention across its entire claimed scope without exercising inventive ingenuity beyond ordinary skill. The adoption of this judicial standard signals that Indian courts will scrutinize AI patent claims with the same rigor applied to mechanical and chemical inventions, rejecting any disclosure that delegates problem-solving to the reader.
The reference to this precedent underscores a broader trend in patent jurisprudence: courts are unwilling to grant monopoly rights for ideas that remain unworkable or speculative at the filing date. For AI applicants, this means that early-stage filings lacking experimental validation, performance benchmarks, or reproducible methodologies face elevated rejection risk.
Practical Risk Mitigation Strategies for AI Patent Applicants in 2025
The combined effect of the January 29, 2025, court ruling and the updated CRI Guidelines creates a stringent compliance environment for AI-related patent filings. Applicants must now include detailed disclosures of training data characteristics, model architectures, hyperparameter settings, and validation protocols to satisfy both enablement and best-method requirements. Failure to provide these technical criteria will trigger rejections under Section 10(4)(a), Section 10(4)(b), and Section 3(k), effectively blocking patent protection for the underlying innovation.
In practice, it is observed that successful filings integrate AI components into larger technical systems, demonstrate measurable improvements in computational efficiency or accuracy, and provide sufficient detail for independent replication by a skilled software engineer. Applicants should conduct thorough prior art searches, prepare detailed flowcharts and architecture diagrams, and include experimental results that validate claimed performance gains before filing.
Legal References
The Patents Act, 1970 – Section 10(4)(a) [India]: Every complete specification shall fully and particularly describe the invention and its operation or use and the method by which it is to be performed. This provision requires clear, complete disclosure enabling a person skilled in the art to work the invention without undue experimentation.
The Patents Act, 1970 – Section 10(4)(b) [India]: Every complete specification shall disclose the best method of performing the invention which is known to the applicant and for which he is entitled to claim protection. This ensures that applicants share practical implementation knowledge rather than withholding superior methods.
The Patents Act, 1970 – Section 3(k) [India]: A mathematical or business method or a computer programme per se or algorithms are not inventions under this Act. This exclusion prevents patenting of abstract ideas unless they produce a technical effect in combination with hardware or applied processes.