Introduction
Sending an email to the wrong person is a mistake almost all of us have made or almost made. To solve this, BlackBerry came up with a feature it wanted to patent: colour-code recipients according to their address domains, so that a wrong address stands out to the sender before the message is sent.[1] However, in Blackberry Ltd v. Controller of Patents and Designs,[2] the Delhi High Court rejected the patent. The refusal itself may be correct, as the claimed invention was found to be obvious over the prior art.[3] What warrants attention is the Court’s second holding under Section 3(k) of the Patents Act, 1970: that a wrong-recipient error is not a “technical problem” because “it depends on the person sending the message.”[4] By adopting this reasoning, the Court has effectively required a technical problem to be universal. This post argues that this requirement finds no support in the statute, the case law or the Patent Office’s own guidelines. It then examines what this requirement puts at risk and how the Court could have reached the same result on firmer ground.
The Case
BlackBerry’s 2008 application claimed a method of colour-coding recipient names in a message on the basis of an address characteristic such as the host or domain name.[5] The Controller refused it on two grounds. First, the invention lacked an inventive step over three prior-art documents (D1, D2 and D3).[6] Second, it was a computer program per se under Section 3(k), because the problem it addressed was administrative rather than technical.[7]
BlackBerry appealed & argued that the feature made a technical contribution, as Ferid Allani[8] requires. It relied on the 2013 CRI Guidelines,[9] which listed an “improved user interface” as a technical effect, and on the Court’s earlier decision in 2024 allowing a BlackBerry appeal under Section 3(k).[10] It also pointed out that substantially the same invention had been granted in Europe and the United States.[11] The Controller replied that the 2013 Guidelines had since been superseded, and that a technical problem is “universal in nature” and does not change from person to person.[12]
The Court sided with the Controller on both grounds. On inventive step, it held that the prior art together rendered the invention obvious.[13] On Section 3(k), it held that a computer-implemented invention must show a technical effect “beyond the abstract logic of a program,”[14] and that any gain in reliability must be “intrinsic to the computer that is an improvement in the machine itself.”[15] It then held that a wrong-recipient error “cannot be considered a technical problem as it depends on the person sending the message”[16] and dismissed the appeal.
Either ground alone would have sustained the refusal, so the Section 3(k) holding stands on its own and can be cited in future refusals.
The Universality Requirement for Technicality
At its core, the Section 3(k) ruling hinges on one premise: that for a solution to count as technical, it has to be identical for everyone.
Yet the Court didn’t come up with this idea on its own. It inherited it from the Controller’s initial order, which used noise in communication systems to argue that real technical problems are universal.[17] That argument was pushed again on appeal,[18] and the bench adopted it, concluding that the original order “rightly states that a technical problem is universal in nature”.[19] However, the judgment doesn’t cite any authority to support this conclusion.
That lack of backing makes sense, because the requirement doesn’t exist in Indian patent law. Section 3(k) excludes “a computer programme per se,” without saying a word about how widely shared a problem needs to be.[20]
Look at established jurisprudence: Ferid Allani focuses on whether an invention delivers a technical contribution or technical effect.[21] The Delhi High Court’s Ab Initio decision clarifies that this effect isn’t limited to hardware tweaks or “nuts and bolts.”[22] The 2025 CRI Guidelines take the same approach, asking concrete questions about functionality: Does the software speed up processing? Does it save memory, improve data transmission, or control a physical device?[23]
All of these authorities focus on what the invention does, not who happens to face the problem. By making universality a requirement, the Court imposed a new criteria for inventors to fulfill.
Why Universality Fails
The problem with the “universality test” is that it focuses on the wrong thing. To figure out if an invention is truly technical, we should be looking at the solution it offers, not how many people experience the problem. Previous cases like Ab Initio and Ferid Allani got this right: they asked whether the invention improves the machine or enhances its processing power.[24] The universality test ignores what the invention does to the computer and instead focuses on who faces the problem. That isn’t what the law asks us to evaluate.
The test doesn’t even hold up when you look at its own examples. The Controller pointed to noise in a communication system as a classic universal problem.[25] But such noise isn’t the same for everyone. It changes depending on your location, your device, and how many other people are online at that exact moment. If a technical problem has to affect every user in the exact same way to count as “universal,” almost nothing would qualify; not even the Controller’s own example.
More importantly, this new test contradicts the Court’s own recent decisions. In 2024, the Court allowed a BlackBerry application for a feature that selected media based on highly personal user habits, like how many times someone played a specific song.[26] The Court ruled that because this increased the device’s capabilities, it was not barred by Section 3(k).[27] A user’s taste in music is incredibly subjective, just as personal as choosing who to send a message to. BlackBerry specifically pointed to this case, but the Court completely ignored it.[28]
To be clear, this doesn’t mean every minor tweak to a user interface should be patented. A real technical effect has to be more than just a standard UI update,[29] and something simple like colour-coding recipient names probably doesn’t make the cut. But the Court already had a perfectly good reason to reject that specific patent: it produced no “improvement in the machine itself.”[30] That reason alone was enough to settle the case. By throwing the “universality requirement” into the mix, the Court just created a new hurdle that will cause uncertainty in future disputes.
What the Test Puts at Risk
The 2025 CRI Guidelines do not exclude display-related inventions altogether. They exclude inventions whose main function is presenting data, but not the ones providing a technical solution to a problem in the way such data is handled or presented.[31] This leaves space for interface inventions that change how a system processes or displays information. A universality requirement would close that door. Many inventions solve problems that differ from user to user, including screen readers and adaptive displays for users with low vision, interlocks that prevent operator error, and warnings that flag phishing or fraudulent transactions. Under the Court’s reasoning, each of these could fail at the threshold, before anyone asks whether the solution is technical.
The Better Route: Section 3(n) and Inventive Step
The Court could have reached the same result without adopting a new test. The Patents Act already excludes presentation of information under Section 3(n).[32] Colour-coding recipient names is a way of presenting information, yet the judgment never mentions Section 3(n). Part of the reason may lie in the Patent Office’s own guidelines. The 2025 CRI Guidelines treat inventions whose purpose is information management and presentation as falling under Section 3(k) rather than Section 3(n).[33] This miscategorisation allowed a presentation objection to be reframed as a new Section 3(k) test.
Either the feature is a presentation of information under Section 3(n), or the claims are obvious because prior art already disclosed differentiating messages by colour according to the recipient.[34] The Court accepted the second route.[35] Neither route, however, required a universality condition.
European practice shows the line can be drawn this way. The EPC also excludes “presentations of information,” but only “as such.”[36] The Boards of Appeal treat presentation as technical only where it interacts with technical features to produce a technical effect.[37] That is an effect-based test with no universality requirement. Notably, BlackBerry secured a European patent for substantially the same invention.[38]
Conclusion
Blackberry may well have reached the right result. The claims were obvious over the prior art, and the feature produced no improvement in the machine itself.[39] However, the Court’s additional holding that a problem is not technical if it varies from person to person has no basis in Section 3(k), in Ferid Allani or Ab Initio, or in the Patent Office’s own guidelines. The Court’s own finding that the invention did not improve the machine was enough. The case also leaves a larger question open: whether Section 3(n) still does any independent work in Indian patent law, or whether Section 3(k) has quietly absorbed it.
[1]Blackberry Ltd v. Controller of Patents & Designs, C.A.(COMM.IPD-PAT) 14/2022, Paras 7.8, 29, 35 (Del HC, 30 April 2026).
[2]Ibid.
[3]Ibid, Paras 31, 32.
[4]Ibid, Para 46.
[5]Ibid, Paras 1–2, 28, 33 (Patent Application No. 1976/DEL/2008, claiming priority from EP 07117003.9).
[6]Ibid, Paras 3, 15.
[7]Ibid, Paras 33–35.
[8]Ferid Allani v. Union of India & Ors., W.P.(C) 7/2014 (Del HC, 12 December 2019), 2019 SCC OnLine Del 11867.
[9]Supra Note 1, Para 8.5.
[10]Ibid, Para 7.23(b); BlackBerry Ltd v. Controller of Patents & Designs, C.A.(COMM.IPD-PAT) 318/2022, Para 44 (Del HC, 30 August 2024).
[11]Supra Note 1, Para 7.13.
[12]Ibid, Paras 8.3, 8.5.
[13]Ibid, Paras 31–32.
[14]Ibid, Para 38.
[15]Ibid, Para 45.
[16]Ibid, Para 46.
[17]Ibid, Paras 34–35.
[18]Ibid, Paras 8.3, 8.4.
[19]Ibid, Paras 46, 47.
[20]Section 3(k), The Patents Act, 1970.
[21]Supra Note 8.
[22]Ab Initio Technology LLC v. Assistant Controller of Patents & Designs, C.A.(COMM.IPD-PAT) 26/2021 & 447/2022, Para 38 (Del HC, 30 July 2024).
[23]CGPDTM, Guidelines for Examination of Computer Related Inventions (CRIs), 2025, Example 7 (Public Notice No. CG/F/CGPDTM/DL-07/919, 29 July 2025).
[24]Supra Note 22, Para 38; Supra Note 8.
[25]Supra Note 1, Para 35.
[26]Supra Note 10, Paras 27–28.
[27]Ibid, Paras 34, 44.
[28]Supra Note 1, Para 7.23(b).
[29]Supra Note 22, Para 38.
[30]Supra Note 1, Para 45.
[31]Supra Note 23, Table 2.
[32]Section 3(n), The Patents Act, 1970.
[33]Supra Note 23, Example 7.
[34]Supra Note 1, Paras 24–26.
[35]Ibid, Paras 31–32.
[36]Convention on the Grant of European Patents art. 52(2)(d), 52(3), Oct. 5, 1973.
[37]European Patent Office, Case Law of the Boards of Appeal § I.A.2.6 (10th ed. 2022) (citing T 1143/06).
[38]EP 2275980 B1 (granted 13 July 2016); see Supra Note 1, Para 7.13.
[39]Supra Note 1, Paras 31–32, 45.
Authored by: Mr. Ashutosh Jha, NALSAR University of Law, Hyderabad