A patent applicant is entitled to fight the case it is told it must meet — not a different one sprung in the refusal order. In Navya Network Inc. v. Assistant Controller of Patents and Designs [2026:BHC-OS:9481], the Bombay High Court set aside a Section 3(k) refusal precisely because the Patent Office moved the goalposts between the hearing notice and the final order.
From “computer program” to “self-learned ontology”
The application (no. 2068/MUMNP/2014, “Medical Research Retrieval Engine”) claimed a system and method for retrieving medical research documents using predefined relational expressions mapped to an ontological hierarchy derived from medical literature. It was refused for lack of inventive step under Section 2(1)(ja) and as non-patentable under Section 3(k), as an algorithm and computer program per se. But the hearing notice had objected that the claims lacked hardware features and were a computer program; the impugned order shifted to a new theory — that the invention was a non-technical “self-learned ontology” algorithm. Navya also argued the invention solved a technical problem (inefficient retrieval) with a hardware-implemented solution improving search efficiency and reducing computational load, and that the Controller had conflated inventive step with Section 3(k).
Two familiar failings
The Court found two now-familiar deficiencies. First, there was no structured inventive-step reasoning: the order neither identified the inventive concept nor compared the claimed features with the prior art, nor explained how a skilled person would get from one to the other — a clear departure from the five-step test in F. Hoffmann-La Roche v. Cipla and Tapas Chatterjee v. Controller. Second, the Section 3(k) finding travelled beyond the hearing notice; introducing a new line of exclusion at the order stage, without a chance to rebut, breached natural justice. And, the Court added, even a meritorious Section 3(k) objection could not cure an otherwise flawed order. Remand followed.
Why it matters
Navya Network is a useful CRI (computer-related invention) data point on two fronts. Procedurally, it reinforces — alongside this blog’s note on Wirtgen — that the case put at the hearing is the case that must be decided. Substantively, it flags the recurring error of collapsing the inventive-step enquiry into Section 3(k): whether a claim is a computer program “per se,” and whether it is obvious, are different questions, and a technical effect (here, retrieval efficiency and reduced computational load) is central to both the eligibility and the inventiveness analysis. For software-adjacent applicants, the order is a reminder to anchor claims in a concrete technical effect and a hardware implementation — and to hold the Office to the objections it actually raised.
Sources & further reading: Section 3(k); Section 2(1)(ja); F. Hoffmann-La Roche v. Cipla.
Educational note: This case summary is general information, not legal advice.