The Calcutta High Court recently upheld the Controller’s rejection of a patent application based on the old 2017 CRI Guidelines. In examining the order, Shailraj Jhalnia writes about the nature of these guidelines and whether the Controller and the Court’s reliance on the older version was justified. Shailraj is a third year law student pursuing B.A. LL.B. from National Law School of India University, Bangalore, with a keen interest in IP Law, Arbitration and Criminal Law. His previous posts can be accessed here.
Figuring Out the Correct CRI Guidelines by Looking at Google LLC v The Controller of Patents
By Shailraj Jhalnia
Section 3(k) of the Indian Patents Act specifically bars “computer programme per se” (and algorithms) from being patented. Over the years, the Patent Office and courts have fashioned an exception for patentability if a software-based invention claimed has a technical effect or improvement. The CRI Examination Guidelines (2017, 2025) and recent case law in Delhi HC, for instance, Raytheon, Microsoft, attempt to delineate this “technical effect” carve-out. Despite this, inconsistency prevails. For instance, in Google LLC v The Controller of Patents (Calcutta HC) – decided on 6th August, 2025. Its a case of a 2014 filed application where the court upheld rejection under S.3(k) without reference to the newer 2025 Guidelines, falling back instead on earlier benchmarks which is rightly held as it is an appeal from an order of 2020 – much earlier than the 2025 guidelines. So now this post identifies an interesting nub that if CRI Guidelines are nothing more than interpretive aids (and not law), should appellate review in 2025 take recourse to the updated Guidelines as the “new lens,” or use the version that held field at the first examination?
Nature of the Guidelines
The IPO’s CRI Guidelines are obviously non-binding interpretative tools, and not legislative tools. The 2017 Guidelines say:
“These guidelines are not rule-making. In the event of any inconsistency, the Act and Rules shall override these guidelines. The guidelines shall be subject to change from time to time as per interpretation by Courts, amendments in the statute, and useful suggestions by stakeholders.”
The 2025 Guidelines also emphasize that they are examiners’ tools, not new law. The Guidelines affirm that statutory provisions take precedence and that guidelines are updated from time to time. In short, the Guidelines themselves describe an “interpretive” role, instead of a binding rule-making function.
The Hierarchy of Legal Authority – Precedent v Guidelines
Under Indian law, domestic patent guidelines for computer-related inventions are non-binding and merely “interpretive aids” under the Act and case law. Prof. Basheer & Prashant Reddy note that IPO manuals “do not have the force of law” and are merely “explanatory and recommendatory”. Likewise, Aparajita Lath sees that guidelines “are not law” and have “no authority to change the law”. In reality, any material patent policy shift (eg, Section 3(k) eligibility) must be legislated by Parliament or the courts. Guidelines can elucidate or administer the statute for examiners, but not alter it or override binding precedents. Accordingly, a High Court is obliged to apply settled case law even if that is inconsistent with IPO guidance. In effect, only Parliament and the courts define patent law; IPO guidelines cannot alter that framework.
Arguments Against Retroactive Application of the 2025 Guidelines
Applying the 2025 Guidelines to appeals of applications determined under the 2017 regime threatens unfairness and uncertainty. As said in this post by Swaraj where there has been back and forth on the position of CRI guidelines with each edition. So the 2025 guidelines are wide apart from the 2017 one, which makes this argument essential, where there is no level playing field for an applicant adjudicated by 2017 and the other with 2025 guidelines having an upper hand. Patent prosecution is highly susceptible to the expertise of the applicant to anticipate patentability criteria. If 2017/18 examiners rejected claims on the basis of stricter interpretation (e.g., innovative hardware), it appears unfair to let a subsequent applicant gain an advantage on a less strict criterion on appeal. Successful appellants may apparently reap a windfall denied to others, similarly positioned but non-appellants. This “moving target” frustrates legitimate expectations.
The Guidelines themselves suggest future operation. To use a specific example, the 2017 CRI Guidelines say that they apply “with immediate effect” on publication – i.e., they apply from then onwards (though with no suggestion of retroactive effect). Furthermore, by their own terms, they are subject to future amendment. That is, the authors did not intend every revision to go back and disturb completed examinations.
Reconstruction of the rules on appeal would unfairly punish those who had acted in good faith on the previous norm. Section 117A appeals are to review examiners’ decisions, not re-do the rules of engagement later.
The European Patent Office updates its examination Guidelines every year, but applies them ahead of time. Pending applications continue to be subject to the law in force at the time of examination. Appeals (to EPO Boards) employ established case law to maintain consistency. Interestingly enough, EPO boards assert unequivocally that the Guidelines are not binding on appeal. Boards employ fresh case law even where it contradicts earlier Guidelines. Considering all these factors, these concerns caution against retroactive application of the 2025 CRI guidelines.
Arguments For Applying the 2025 Guidelines on Appeal
Conversely, applying strictly earlier standards of appeal can be error-maintaining. Guidelines are interpretive tools, and accordingly, in a de novo appeal, one would utilize the superior present vision of the law. Appeal review is not simply the continuation of the examiner; the High Court (or Controller) reexamines questions of law afresh. If later jurisprudence has established the scope of S.3(k), appeals would be negligent not to take it into account.
The Guidelines are “only a lens” by which to look at the statute. If that lens has been updated to remove misconceptions, why cling to the blurriest vision? The 2025 Guidelines formally embrace new Delhi HC precedents (Raytheon) that remove the novel-hardware requirement. Appeals based solely on the 2017 lens would be in danger of using standards that courts have held legally unsupported.
The 2025 Guidelines codify Delhi HC’s practice (e.g., prioritizing technical effect/contribution over hardware). If an appeal ignored this trend, the court could sub silentio overrule binding precedent. Indeed, the court “categorically barred” any new-hardware test in Raytheon, and Microsoft demanded an apparent, plausible technical effect rather than mere computing. Appeals ignoring these holdings would undermine the uniform application of patent law.
Adopting the newer, liberal approach on appeal can spur technology development. The updated Guidelines aim to ease excessively high patent hurdles for computer program inventions – in line with India’s aspirations for a digital economy. Implementing the 2025 vision on appeal would rescue meritorious CRI inventions needlessly killed by old tests, hence, software and AI R&D.
The Problem at Hand
The stakes are illustrated by the Google ruling of the Calcutta High Court. The 2014 patent application had been rejected in 2020 under S.3(k) on the grounds that there was no “technical effect.” Google argued in the 2025 appeal that the invention (labelling of contact data location) offers a technical advantage to an abstract algorithm. The Controller replied with the 2017 test (three-pronged CRI test, encompassing novel hardware) and sought to dismiss.
The High Court upheld the rejection. It found the entire invention works on standard software, “without reciting any novel technical implementation, hardware integration, or measurable technical contribution”. The described process was found to be a mere routine data-processing abstract. The court specifically found the examiner’s determination that “no technical effect” to be correct and “does not justify any interference”.
Ironically, the ruling cites the exclusion of S.3(k) and the test (2017 and 2025 versions verbatim) by the CRI Guidelines, but never adopts the Assistant Controller’s formulation: no patent if all the steps are done in generic computer memory. Instead, the Court de facto ratified the Assistant Controller’s formulation.
By refusing to extend Raytheon’s reasoning, Google maintains the existing regime. It is a “lost opportunity” to align with the more liberal approach of the Delhi HC. The case raises the vital question that this blog tries to answer: Is an appellate court in 2025 bound by an earlier interpretative precedent merely because the prosecution of the application was of older vintage? Google contends it is, but not without criticism.
Policy Considerations
Both are arguments based on value. Prospective application of guidelines (as opposed to retrospective application) is a respect of certainty and equity: applicants can have recourse to settled rules, and agencies will not retrospectively move the goalposts. It preserves the existing prosecution expectations. There would be grounds that it would not decrease the backlog of cases if the 2025 guidelines do not come into force; however, in the long-term scenario, it would keep the patent regime in line and would not leave inventors and companies in chaos. Therefore, I believe Google’s judgment is good in law for the aforementioned reasons.
Conclusion
The 2017 v 2025 CRI Guidelines controversy is a manifestation of a deeper tension between correctness and predictability. There is no one correct answer. Equity does not require punishing applicants by midstream changes; correctness requires applying the law as understood currently. Patent applicants and examiners are left uncertain without clear statutory direction on how S.3(k) is to change. In practice, it might be reasonable to state that the IPO will apply new Guidelines prospectively explicitly, but courts of appeal will decide S.3(k) claims according to current case law (e.g., Raytheon and Microsoft principles). Until then, CRI results will continue to depend on whether one perceives the Guidelines as a mere examiner tool or a binding interpretive policy.