One of the most persistently litigated grounds for rejecting Indian patent applications in the software and computational fields is Section 3(k) of the Patents Act, 1970, which excludes 'a mathematical or business method, a computer programme per se, or algorithms' — and, closely related to that, what examiners often term a 'mental act': a method that could, in principle, be carried out by a person thinking it through rather than by a machine performing a genuine technical operation. The Delhi High Court, in a ruling delivered by Justice Tushar Rao Gedela, has now stepped in to give applicants and examiners a workable line between the two.
The Court's framework turns on whether the claimed method produces a technical effect that goes beyond what could be achieved by human cognition alone. A method is not disqualified simply because a human being could, hypothetically, follow the same logical steps in their head — the relevant question is whether the claim, taken as a whole, discloses a technical implementation that interacts with hardware, data structures, or physical processes in a way that produces a concrete, verifiable technical outcome. Where the claim does no more than describe an abstract sequence of reasoning or decision-making — however sophisticated — it remains an unpatentable mental act, regardless of the field it is dressed up in.
The ruling sits within a broader, decades-long international conversation about where the line between 'idea' and 'invention' should fall in computer-implemented inventions — comparable in spirit to the 'technical contribution' tests developed by the European Patent Office and the UK courts, though rooted firmly in India's own statutory text and the Patent Office's Computer Related Inventions Guidelines. What the Delhi High Court has added is a degree of doctrinal clarity that has been missing at the examination stage, where Section 3(k) objections have often been raised (and litigated) somewhat inconsistently across applications with materially similar claim structures.
For applicants filing software, AI, and data-processing patents in India, the practical lesson is in claim drafting rather than invention itself: applications should foreground the technical means by which a result is achieved — the specific data processing architecture, hardware interaction, or measurable technical improvement — rather than describing the invention primarily in terms of the logical or business outcome it produces. Claims that read as a set of mental steps dressed in technical vocabulary remain the most vulnerable to a Section 3(k) objection even after this ruling.