An Introduction to Software Patents
Software Patents are an aspect of Indian patent law that has proved to be difficult to regulate since the turn of the century, due to the exponential growth in computer applications and technological innovations. Section 3(k) of the Patents Act, 1970 was amended in 2002 to limit non-patentability of Computer-Related Inventions (CRIs) to only algorithms, mathematical and business methods, and computer programmes “per se” to account for these developments. These two words, “per se”, lacking any sort of statutory definition, have since found themselves at the center of a divisive storm of opinions surrounding software patents, which have themselves gained a tacit approval from the Indian Patent Office and the Courts, despite being non-patentable as per the letter of the law.
SFLC.in has always rejected the broadening of scope of Section 3(k) to permit the patenting of computer programmes that have “technical effect”. We have written extensively on this, our stance remains firm: Say No to Software Patents! In 2023, we substantiated this stance with data, through our report titled “Software Patents in India: Law and Practice” (“2023 Report”) where we recorded that a significant majority (around 80%) of all software patents granted in India each year since 2020 were granted to foreign applicants, much to the detriment of domestic software developers and the FOSS community in the country. Through this blog series, we aim to analyse the current legal regime for software patents in India, and describe the impacts of the same.
Part I of the series is an examination of the “Guidelines for Examination of Computer Related Inventions (CRIs), 2025”, (“2025 Guidelines“) and the substantive changes made therein. Part II of the series will examine the judicial decisions surrounding software patents, and the inconsistent and excessive number of disparate judicial tests that have been established for the same. Part III of this series will provide a detailed look at the specific impacts of software patent grants on the domestic software field, as well as a comparative analysis of software patent regulation globally.
Legislative Intent and History Behind the Exclusion
Before examining the CRI Guidelines, it is pertinent to briefly describe the statutory landscape surrounding software patents in India, and its legislative history. When the Patents Act, 1970 was first passed, Section 3, which describes what are not considered patentable inventions, listed items from (a) to (i), and did not list computer programmes, algorithms, or mathematical and business methods. However, Section 2(1)(j) of the original Act had a narrow definition for “inventions” that emphasised on manufacturing, leaving software out of the purview of the Act.
In 2002, the Patents Act 1970 was amended to be more compliant with The Agreement on Trade-Related Intellectual Property Rights (TRIPS Agreement). Section 2(1)(j) as amended by the Patents (Amendment) Act, 2002 now defined “invention” as “a new product or process involving an inventive step and capable of industrial application” and introduced product patents in India. The other relevant change made by the 2002 Amendment was in Section 3, where sub-sections (j) to (p) were added. As per the newly-added Section 3(k), “a mathematical or business method or a computer programme per se or algorithms” were not inventions under the Act and therefore were not patentable. The words “per se”, inserted to qualify computer programmes alone, meant computer programmes “as such” were not to be granted patent. The Joint Parliamentary Committee on the Patents (Second Amendment) Bill (2001) justified this insertion, stating “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.” As we observed in the 2023 Report, this insertion allowed patents to be granted for computer programmes including other products or processes such as hardware.
Section 3(k) was further amended once more in 2004 through the Patents (Amendment) Ordinance, 2004 for a brief period. The amended provision was bifurcated, with Section 3(k) dealing exclusively with “computer programme per se other than its technical application to industry or a combination with hardware,” and Section 3(ka) dealing with “a mathematical method or a business method or algorithms.” The added qualification of technical application to industry, being vague and overbroad, resulted in strong protests regarding the effect that software patents would have on software. As we noted in the 2023 Report,“it would be difficult to design a software with the multitude of patents granted on software and further as software programmes are only expressions of algorithms in human readable form, they are not patentable subject matter.” In particular, the Free Software community campaigned against the provision and garnered wide non-partisan support, asking the crucial question, “Why was such an amendment required when the Indian patent regime was already TRIPS-compliant in software patents?”
Although the provision was included in the Patents (Amendment) Bill, 2005, it was removed after discussion because it would “give rise to monopoly of multinationals”, and as then Lok Sabha MP Sri Rupchan Pal explained, the Bill was introduced to provide patents to embedded software but the provision was opposed as it would not benefit Indian professionals. The Patents (Amendment) Ordinance, 2004 was subsequently repealed by the Patents (Amendment) Act, 2005, and the language of the 2002 Amendment was retained in Section 3(k) and has remained as such to this day.
Thus, it can be concluded that the intent of the current law is to keep computer programmes per se out of patentable subject matter, that technical effect requirements are antithetical to this intent, and that even embedded software are not to be permitted to be patented. This conclusion forms the basis for our advocacy against the increasing approval of software patents that we have observed in the past decade.
The CRI Guidelines Before 2025
In July 2025, the Indian Patent Office released the 2025 Guidelines aiming to provide guidelines for the examination of patent applications in the field and to “bring out clarity in terms of exclusions expected under section 3(k)” to “efficiently and effectively” examine CRI patent applications. This release marks the fourth set of full guidelines published on CRIs by the Patent Office, following the 2015, 2016 and 2017 Guidelines.
SFLC.in analysed the major impacts of each of these guidelines in our 2023 Report, which can be accessed here. We found that the 2015 Guidelines allowed software patents except for cases where the “substance of the claim” falls outside the excluded categories. The aim was to ensure the exclusion is not avoided ”merely by camouflaging the substance of the claim by its wording,” putting an end to the allowing claims that camouflage software patents by clubbing them with hardware. The 2016 CRI Guidelines added additional obligations to any Computer-Related Invention claims, in no small part due to the three-part test that required the claim to be examined in substance, and denied unless the actual contribution extended beyond a computer programme or algorithm, and involved computer programme combined with novel hardware. However, the 2017 CRI Guidelines immediately worsened things, treating software patents the same as other general inventions, and removing the specific three-part test. As a result, there was a significant increase in the number of software patents being granted, from 331 the previous year to 842 in 2017.
The 2025 Guidelines, passed in light of the complex and pervasive developments in Information and Communication Technology and Computer Science such as “advancements in artificial intelligence (AI), blockchain technology, quantum computing, cloud computing and the Internet of Things (IoT),” therefore also merit examination along the same lines.
The 2025 CRI Guidelines
The 2025 Guidelines was released on 29th July, 2025 after two sets of drafts, released in March and June respectively. The stated objective of the 2025 update to the Guidelines is to account for the advancements in various complex technologies such as AI, IoT, blockchain, quantum computing, etc. In contrast to the previous guidelines, the 2025 Guidelines dedicate an entire portion to the legal provisions and jurisprudence relating to CRIs, relying on judicial pronouncements from various Courts in the country to both inform the current examination standards and to allow the examiners to seek guidance from the law when accepting or rejecting CRI applications. However, there are some broad concerns to be found in the 2025 Guidelines
1. Over-reliance on foreign tests: In Para 3.5 of the Guidelines, there is extensive reference to the various tests developed by various Courts, including Ferid Allani vs. Union Of India & Ors [W.P.(C) 7/2014 & CM APPL. 40736/2019, Delhi High Court] which influenced multiple subsequent decisions and permitted software patents for computer programmes per se if it demonstrated a “technical effect” and “technical contribution”. While a detailed analysis of this judgement, and the other judgements relied on in the Guidelines, will be provided in Part II of this series, it is worth noting that the judgement in Ferid Allani relied extensively on foreign jurisprudence and tests from “across the world” to reach its verdict, and omitted to consider the detailed history of Section 3(k).
The other judgements mentioned in the CRI Guidelines also exhibit the same pattern, and rely more on judicial pronouncements and examination standards from other countries, where the law is significantly different from the bar on patentability found in the Patents Act, 1970. For example, EU Patent Convention, Article 52(2)(c) does not exclude algorithms from patentability, and computer programs are explicitly patentable if they have “technical character” that produces a “further technical effect.” Similarly, UK Patents Act 1977, passed in light of the European Convention and judgements of the European Patent Office, also allows for software patents if it meets the “technical contribution” as per the AT&T signposts, a judicial test that follows the parity between the English and European patent regimes first established intentionally in In The Matter Of Application No. 9204959.2 By Fujitsu Limited [1997] EWCA Civ 1174 (6th March, 1997). The Indian Patents Act, 1970 is markedly different from the above laws, and its application ought not to be influenced by foreign jurisprudence that ignores the long and complex legislative history of the bar on software patents.
To illustrate this point, one has to look no further than the definition of “per se” in the 2025 Guidelines, which uses the definition in Para 25 of the judgement in Microsoft Technology Licensing LLC vs Assistant Controller of Patents and Designs (T) CMA (PT) No.49 of 2023 [OA/36/2020/PT/CHN] (3 July 2024, Madras High Court). However, what is truly illuminative about this judgement lies in Para. 29, where the Hon’ble Madras High Court concludes its examination of the legislative history of Section 3(k) stating, in reference to the Patents Act (Amendment) Bill, 2005, that “software related patent applications cannot be measured solely against the benchmarks of technical application to industry or combination with hardware. Parliament’s intention was clearly not to lower the bar to the extent indicated in the above mentioned Bill.” Despite this acknowledgement of the legislative intent, the Hon’ble Court went on to examine the UK and EU Patent jurisprudence owing to the presence of “per se” and “as such” in these foreign jurisdictions, and relied on Ferid Allani as an endorsement of these foreign tests. Resultantly, the definition of “per se” in Section 3(k) has become congruent with the definition in foreign statutes, without consideration to the legislative intent behind the provision.
2. Lack of uniformity and consistency in tests for examination: Further, there is no uniform test to be used for determining whether a computer programme is patentable subject matter, as there is extreme fragmentation and divergence in judicial opinion, a pattern observed by SpicyIP as well. As Swaraj Paul Barooah rightly puts, “In the draft guidelines, some excerpts from some cases are mentioned. However, why are those mentioned in the sections they are? What about those cases is authoritative in terms of guidance that can be taken? What about the ones, from equal judicial authority, that are not mentioned? Why are some being chosen over others?”
This fragmentation results in selective application of judicially established tests and selective reliance on judicial opinion, something evidenced in the 2025 Guidelines themselves. While the judgements listed in Para 3.5 rely on a range of tests from “technical effect” to “technical solution to a technical problem” to “beyond usual user interface,” the steps for assessing the claimed inventions listed in Para 4.5.4.1.3. only requires the Patent Examiner to determine “whether the identified technicality results in a technical effect which is beyond a mere incidental effect,” a language not found in any of the referenced judgements. The same procedure is reiterated in Para 4.5.5.4, in flow chart format. Additionally, in OpentvInc vs The Controller Of Patents And Designs [C.A. (COMM.IPD-PAT) 14/2021], it was explicitly held that the “bar in India to grant of business method patents has to be read as an absolute bar without analysing issues relating to technical effect, implementation, technical advancement or technical contribution.” Despite referencing this exact holding in the 2025 Guidelines, the assessment steps in Para 4.5.2.1 allow for exclusions of business methods if it is “is technical improvement/ solution to an underlying system or process, aimed at refining operational framework or infrastructure, and using business context only as a constraint to define the scope of the invention.”
3. Already stifles domestic patent ownership, which will intensify due the AI boom: The current Software Patents regime, and the 2025 CRI Guidelines which regulate their approval, has increased relevance in the age of AI, when the boundaries between computer programme and algorithm have started to blur. In 2021, the One Hundred and Sixty First Report of the Parliamentary Standing Committee on “Review of the Intellectual Property Rights Regime in India” was presented before the Rajya Sabha. In Para 8.3. of this report, it is emphatically stated that “As per Section 3(k) of the Indian Patent Act,1970, a mathematical or a business method or a computer programme or algorithms run by Artificial Intelligence are not patentable.” However, the report also calls for a review of the provision due to the same reason, further recommending in Para 8.7. that “the approach in linking the mathematical methods or algorithms to a tangible technical device or a practical application should be adopted in India for facilitating their patents as being done in E.U. and U.S.” Not only would such a development negatively impact millions of software developers across the country, who would then have to pay in order to utilise AI algorithms which have been patented, but it is also a stark departure from the legislative intent of the provision.
At this juncture, it is pertinent to examine the Statement of Objects and Reasons attached to the Patents Act, 1970, which states plainly and clearly that the main purpose of the Act is “stimulating inventions among Indians and encouraging the development and exploitation of new inventions for industrial progress in this country…” and to be “…a more comprehensive law so as to ensure that patent rights are not worked to the detriment of the consumer or to the prejudice of the trade or the Industrial development of the country…”
As we noted in our 2023 Report, a majority of software patents in India are already granted to foreign entities, notably Big Tech companies like Microsoft, Google, and Oracle. If the regulatory standards governing software patents remain lax as they are, we expect similar trends to occur in the AI patents sphere as well, stifling domestic progress in the field as well as open source AI movements in the country.
Conclusion
Despite the clear and well-documented legislative history demonstrating the Parliament’s consistent intent to exclude computer programmes per se, algorithms, and business methods from patentable subject matter, the 2025 Guidelines utilise selectively chosen judicial pronouncements and foreign patent standards that are incompatible with the Indian patent framework, to turn a statutory exclusion into a discretionary exercise for patent examiners. Allowing software patents, whether directly or indirectly, contradicts both the text and purpose of Section 3(k) and risks enabling monopolies, harming progress and domestic software developers.
