The issue of Computer Related Inventions (CRI) in India has popped up again, with Delhi High Court presently hearing an appeal against an order of the Patent Office rejecting an application filed by Microsoft, titled ‘Reversible 2-Dimensional Pre-/Post- Filtering For Lapped Biorthogonal Transform’, under Section 3(k). SpicyIP intern Sukarm Sharma discusses this ongoing dispute, in light of the CRI 2016 guidelines and the Ferid Allani order. Sukarm is an undergraduate law student at NLSIU, Bangalore.
Microsoft v. Asst. Controller of Patents and Designs: Revisting the CRI Guidelines and Ferid Allani
Sukarm Sharma
Software patenting remains one of the most contentious issues in patent law jurisprudence, with an ever ping-ponging legal position. This is brought out about by the ever-oscillating interpretation of the clause that software cannot be patented ‘per se’. Swaraj Barooah has done an excellent job of summarizing these developments in his piece here. This post, however, focuses on the grounds for appeal to the 2019 decision of Microsoft Technologies Licensing (“Microsoft Technologies”) by the Bombay Patent Office (pdf), which is now up for appeal at the Delhi High Court. The appeal is currently in the process of being heard, with the latest hearing on 6 March 2023 (pdf). Consequently, this post grapples with questions relating to the legal nature of Computer Related Inventions (“CRI”) guidelines, legislative intent and other related issues potentially relevant to the appeal.
The Bloopers in the Bombay Patent Office’s Decision
The appeal to the decision in Microsoft Technologies could have serious implications on software patenting law in India. In short, the Bombay Patent Office rejected Microsoft’s request to patent their software titled “REVERSIBLE 2-DIMENSIONAL PRE-/POST- FILTERING FOR LAPPED BIORTHOGONAL TRANSFORM” on the grounds of the Section 3(k) exception. To the uninitiated, the Section 3(k) exception simply lays down that “mathematical or business method or a computer programme per se” cannot be patented. However, since this leaves a fair bit of room for interpretation, various CRI guidelines have been released by the Office of the Controller General of Patents, Designs and Trade Marks. These have subsequently undergone several revisions, a succinct timeline of these changes till 2016 is linked here. Notwithstanding their consistently changing nature, they have largely been adhered to while deciding on patent applications.
The pertinence of this to the matter at hand is – that while deciding on whether Microsoft’s technology was eligible for a patent, it relied on CRI’s 2016 guidelines, instead of the 2017 guidelines. This is relevant, because the application of the 2017 guidelines would likely have yielded an entirely different result. The 2016 guidelines were far more restrictive in granting software patents. They laid a clear 3-part test to evaluate software claims, which inter alia outrightly rejected purely software-based patent claims, and required that any combination of software-hardware necessarily have novel hardware to be eligible for a patent. In contrast, the 2017 guidelines were far more relaxed. They removed the three step test, including the requirement for a novel hardware, in effect clarifying that there was no absolute bar on purely software-based patent claims. A comparative appraisal of these guidelines is available here.
Now, the relevance of the Bombay Patent Office’s reliance on the 2016 guidelines rather the 2017 ones should become clear. Microsoft’s application – for a purely software claim, would simply not be eligible under 2016 guidelines, while the 2017 guidelines have no such bar. The Patent Office’s claim, that the “the computer programme in itself is never patentable. If the contribution lies solely in the computer programme, deny the claim” is simply untenable on an application of the 2017 guidelines.
What is the Legal Nature of the CRI Guidelines?
This however, raises an important question – What is the nature of the CRI Guidelines that seem to be rewriting the text of Section 3(k)? It’s clear that they are not any form of legislation, delegated or otherwise. In fact, paragraph 1.5 explicitly mentions that “these guidelines do not constitute rule making” and that in case of any conflict, “the provisions of the Act and Rules will prevail over these guidelines.” Even more perplexingly, the 2017 revisions have been considered by the government to be merely ‘clarificatory’ and not a regime change over the 2016 ones. Since these guidelines have been explicitly clarified to not constitute rule making, it is evident that they are not binding on the court, or even the patent offices. At best, they are tools for interpretation. In that light, would reliance on a 2016 guideline over a 2017 guideline be sufficient grounds for appeal? Strictly on the law, perhaps not, since the guidelines were never a source of law in the first place. In everyday practice, however, the guidelines are relied upon, so the actual decision might be anybody’s guess.
Another interesting aspect is, that since the CRI guidelines clarify that the provisions of the act would prevail over the guidelines, one could argue that the CRI 2017 guidelines are themselves conflicting with the text of Section 3(k), read with legislative intent. The Joint Parliamentary Committee clarified on the rationale for including ‘per se’, in Section 3(k) – “This change has been proposed because sometimes the computer programme may include certain other things, ancillary thereto or developed thereon. The intention here is not to reject them for grant of patent if they are inventions. However, the computer programmes as such are not intended to be granted patent.” This makes it fairly evident, that computer programs by themselves were not intended to be patented under Section 3(k), clarifying an intent akin to the ‘novel hardware’ threshold of the 2016 guidelines. It can very well be argued, that by allowing the patenting of software by itself, the 2017 guidelines conflict with Section 3(k) and therefore become inoperative to the extent of the conflict, bringing us back to the position of the 2016 guidelines.
Concluding Thoughts: Revisiting Ferid Allani
Regardless, another ground for appeal is available to Microsoft through reliance on the Ferid Allani v. Union of India order, passed in 2019, after the rejection of Microsoft’s application to the patent office. It clarified that software by itself, could be patented, provided that it had some ‘technical effect’. Consequently, on precedential grounds, Ferid Allani could form a legitimate basis for appeal. However, it is relevant to note that Allani, similar to the 2017 guidelines, has been criticized for being unfaithful to original intent, since it appears to rewrite the text of Section 3(k) to allow for purely software patenting. The judgement largely relies on policy reasoning, noting that in the digital age, software patents are pertinent to foster innovation, and a failure to recognize the same would be retrograde. While this reasoning may have some merit, it is difficult to accept it when it seems to run counter to the clear text, and the legislative intent. This is evident upon perusing the Parliament’s 161st report on the Review of the Intellectual Property Rights Regime in India. It notes that the current patent regime under Section 3(k) does not adequately protect AI since software patents are not protected, which calls to question the court’s observations that “Innovation in the field of artificial intelligence […] would be based on computer programs, however the same would not become non- patentable inventions – simply for that reason.” While the parliament’s report suggests amending Section 3(k) to permit software patenting, the court seems to mysteriously read that in by itself. It’s also worth noting, that even the policy justification, both by the court and the report, is not a fact beyond question. Easy patentability by itself does not necessarily foster innovation, which requires a careful balance of incentivizing innovation while also avoiding monopolization. This naturally requires careful economic and empirical research, which gets lost in the discourse of more ‘patents=more innovation’. As pointed out in a previous post, the decision also fails to adequately consider the specific facts of the cast. In that light, this appeal might be a useful opportunity to overturn, or modify the position of law set in Allani.