Awards and Recognitions

Registration Is Not Immunity: The Kerala High Court’s Trade Mark Lesson in M. Manuel v. Malabar Gold

By SSR Trademark Team

In M. Manuel v. Malabar Gold Private Ltd.[1], the High Court of Kerala revisited a set of questions that sit at the centre of Indian trade mark practice: what a registration actually protects, when one registered proprietor may sue another, and what separates a winnable infringement claim from a winnable passing off claim. The judgment repays close reading precisely because it resists the easy narrative. The appeal was allowed in part: the plaintiff kept an injunction, but a narrower one, and lost both its passing off finding and its attempt to monopolise a geographical word. Malabar Gold Private Ltd, the Kozhikode jeweller, sued M. Manuel, trading as Malabar Fashion Jewellery in Delhi, for infringement of its registered device mark “MALABAR GOLD” and for passing off. The plaintiff’s registration sits in Class 14 (jewellery), effective 8 March 2002, originally in the name of its predecessor, Malabar Realtors (P) Ltd, and, importantly, subject to an express limitation denying any exclusive right to the word “Malabar.” The defendant, meanwhile, held a registration in Class 35 (advertising and business services), obtained on 17 August 2008 after abandoning an earlier Class 14 application filed under mistaken advice. The trial court decreed the suit on both infringement and passing off. The appeal tested every plank of that decree. The defendant’s headline argument was structural: both sides were registered proprietors, so, relying on Section 28(3) of the Trade Marks Act, 1999 and the Supreme Court’s decision in S. Syed Mohideen[2], no infringement action could lie between them. The Court accepted the principle but denied its application, and the reasoning is the most useful part of the judgment. Section 28(3), the Court held, cannot be read apart from Section 28(1), which confers exclusivity only for the goods or services actually registered. The bar on infringement actions between co-proprietors therefore engages only when two conditions coincide: the marks are identical or nearly resemble each other, and the registrations relate to the same goods or services. Where the rival registrations fall in different classes, one for goods and the other for services, Section 28(3) does not automatically bar an infringement action. S. Syed Mohideen, the Court noted, concerned rival claims over the same category of goods; it did not decide the cross-class situation. The reference order in Abros Sports[3] was, by its own nature, only a question referred for authoritative determination, and the Court declined to treat it as binding precedent. The factual finding sealed it. The defendant admitted, in its written statement and in reply to the caution notice, that it traded in gold jewellery, a Class 14 good, and DW1 conceded that it carried on no Class 35 activity at all. A registration in an irrelevant class, the Court reasoned, cannot protect a jewellery business that falls squarely within the plaintiff’s registered field. Because the defendant had never challenged the plaintiff’s Class 14 registration, the plaintiff’s exclusivity in that class stood protected under Section 28(1) read with Section 29, and infringement was made out. The related jurisdictional arguments under Section 30(2)(e) and Section 124 failed for the same want of pleading: no invalidity defence was raised, no issue was invited, and, per Patel Field Marshal Agencies[4], no prima facie tenability finding was ever sought, so no stay arose. “Malabar” and the Secondary-Meaning Threshold On the geographical objection, the Court found for the appellant. “Malabar” denotes a region of southern India, and geographical, descriptive and laudatory expressions remain open to honest traders unless shown to have acquired a secondary meaning identifying a single source. Two features told against the plaintiff: its own registration expressly disclaimed the word “Malabar,” signalling that no acquired distinctiveness had been recognised at registration, and no consumer surveys, recognition studies or comparable evidence were led to prove that “Malabar” standing alone had come to mean the plaintiff. Turnover and advertising spend, relevant to goodwill in the composite mark, could not by themselves confer a monopoly over the place name. Relying on Parakh Vanijya[5] and a line of Supreme Court authority, the Court held that the plaintiff could claim no exclusivity over “Malabar” in isolation. The Anti-Dissection Rule and Trade Dress That did not end the enquiry. Applying the anti-dissection rule and the tests distilled in Pernod Ricard India[6], the Court compared the marks as wholes. Beyond the shared word, it found substantial similarity in the manner of writing, the visual emphasis on the letters “M” and “R,” the lettering, colour scheme, layout, logo and even the captions: the word “Malabar” housed in a kinked square plank with “GOLD” set boldly beside it. Viewed through the eyes of an average consumer of imperfect recollection, the likelihood of confusion was real, and the defendant offered no honest justification for adopting so close a get-up. Deceptive similarity of the composite mark was therefore affirmed, even though the word itself remained free for use by others. Here the decree unravelled for the plaintiff. Deceptive similarity, the Court stressed, is not passing off. A passing off claimant must prove the classical trinity, namely goodwill, misrepresentation and damage, and must prove it properly. Following Brihan Karan Sugar Syndicate[7], sales and advertising figures must be established through admissible evidence: a statement is not proved merely by being filed. Yet the plaintiff’s turnover and advertising figures were not produced before the trial court at all, surfacing only later as additional evidence on appeal. The trial court had thus found goodwill on the strength of two interested employees, PW1 and PW2, and the plaintiff’s own brochures, with no independent witness: no customer, dealer or certifying accountant. Prior use was made out; misrepresentation and proved goodwill were not. The passing off finding was accordingly set aside. The remaining findings favoured the plaintiff. On prior use, the registration in the predecessor’s name (effective 8 March 2002, claiming user from 19 June 2001), the Section 31 presumption attaching to the recorded assignment, and invoices from 1994 outweighed the defendant’s unsupported claim of use from 1990. On delay, the Court applied Express Bottlers[8]: mere lapse of time is no defence absent proof of conscious, intentional abandonment, and acquiescence needs more than inaction. On jurisdiction, a composite suit for infringement and passing off was maintainable under Section 134(2) of the Trade Marks Act and Section 20 of the Code of Civil Procedure. The parties’ later change of logos did not render the matter infructuous; it went only to moulding relief. For domestic and international brand owners alike, the case is a compact tutorial. First, class discipline is not a formality: India follows the NICE classification, and a registration protects only the field it names, so businesses should file in the classes that match what they actually sell or do. Second, India’s dual system rewards different proofs: statutory infringement flows from a valid registration, while common-law passing off, and India’s prior-user rule, reward genuine use and goodwill. Third, evidence wins cases: goodwill must be proved through admissible material, a point overseas brand owners relying on group figures or in-house testimony too often overlook. Fourth, place names and other descriptive terms can anchor a strong composite brand, but not a monopoly over the word itself. How each of these plays out depends on the marks, the classes and the evidence in any given case, which is where tailored advice earns its place. [1] M. Manuel v. Malabar Gold Private Ltd., RFA No. 7 of 2016, High Court of Kerala at Ernakulam, judgment dated 30 June 2026, neutral citation 2026:KER:46835. [2] S. Syed Mohideen v. P. Sulochana Bai, (2016) 2 SCC 683. [3] Abros Sports International Pvt. Ltd. v. Ashish Bansal and Ors., MANU/DE/3242/2025 (Delhi High Court, reference order). [4] Patel Field Marshal Agencies and Another v. P.M. Diesels Limited and Others, (2018) 2 SCC 112. [5] Parakh Vanijya Private Limited v. Baroma Agro Products and Ors., MANU/SC/0722/2018. [6] Pernod Ricard India Private Limited and Another v. Karanveer Singh Chhabra, 2025 SCC OnLine SC 1701. [7] Brihan Karan Sugar Syndicate Private Limited v. Yashwantrao Mohite Krushna Sahakari Sakhar Karkhana, (2024) 2 SCC 577. [8] Express Bottlers Services Pvt. Limited v. Pepsico Inc., (1991) 11 PTC 296 (Bom) (DB). a Not necessarily. The Court confirmed that the protection given to co-owners of similar marks applies only where the registrations are in the same class, and that a registration in a class a business does not actually operate in will not shield it from a claim founded on someone else’s class. Whether a registration truly matches a business’s activities is a fact-specific question worth reviewing with counsel rather than assuming. a A business can build a protectable composite brand around such a word, but it generally cannot monopolise the word itself unless it proves the word has acquired a secondary meaning pointing only to that business, typically through robust evidence. Turnover and advertising alone are usually not enough. a No. Similarity may establish a likelihood of confusion, but passing off also requires proof of genuine goodwill and damage, established through admissible evidence rather than assertions or the claimant’s own employees’ testimony. In this case, the claim failed largely on proof. a Often it is the other way around. Indian law protects the prior user, whose rights can rank above a later registrant’s. Establishing prior use, however, requires continuous, documented use, such as invoices and business records, rather than mere assertion. a Three things stand out: file in the classes that match the business’s actual goods and services under the NICE system; keep proper, independent evidence of reputation, since Indian courts expect goodwill to be properly proved; and remember that prior use carries real weight in India.

Biological Material Patents in India: Why a Pending NBA Application Should Not Mean Refusal

By SSR Patent Team

There is a quiet but persistent problem in Indian patent prosecution, and Manu Chaudhary v. Controller of Patents and Designs has finally named it. The applicant’s patent (Application No. 201711047431), an oral herbal painkiller composition and its process of preparation, had been refused by the Controller on 22 February 2024. Inventions that draw on Indian biological material must clear the National Biodiversity Authority. That much is settled, and sensible. What has been far less settled is what should happen when the applicant has applied to the NBA in good faith but the approval (which is entirely in the authority’s gift, not the applicant’s) has not issued by the time the office wishes to close the file. Until now, the guidance on this point has been thin, and refusal has too often filled the gap. The Delhi High Court has now held, correctly in our view, that deferral is the correct course. The Biological Diversity Act, 2002 and the traditional-knowledge Guidelines make NBA permission a condition of grant. They do not make its absence, at an interim stage, a ground of refusal. That is not a technicality; it is the whole architecture of the scheme. A condition of grant is satisfied at the point of grant. To convert it into an examination-stage guillotine (refuse now, approval or no approval) is to read into the statute a consequence Parliament did not provide. Justice Singh declined to do so, and rightly. The Court found no provision, in either the statute or the Guidelines, requiring refusal in these circumstances, and none was shown to it. The deeper question the case answers is one of allocation of risk. An applicant can file her NBA application; she cannot compel the authority to decide it by any particular date. To refuse her patent because the NBA has not yet acted is to make her carry a delay she is powerless to prevent. That is neither fair nor supported by the text. The applicant here did more than the bare minimum: she applied, and she put the acknowledgement before the Controller in writing before the order fell. On those facts, the Court held, refusing the application did not reflect a proper exercise of that discretion. Building an IP strategy around Indian biological resources? The biodiversity framework rewards applicants who plan early and punishes those who improvise. Our patents team helps you do the former. Contact S.S. Rana & Co. The same instinct runs through the Court’s treatment of the other grounds. Section 15 was read as a discretion to permit compliance, not a licence to refuse at the first stumble. The objection that the applicant had appeared through the wrong kind of representative (the Respondent’s objection under Sections 128 and 132 of the Patents Act) was treated as a curable defect, to be fixed on remand, not a reason to end the matter. There is a philosophy here worth naming: procedural requirements exist to secure a fair and orderly process, not to defeat otherwise meritorious applications on points that could be put right. It is a principle worth keeping in mind by anyone navigating patent prosecution. It would be a mistake to over-read the decision. The Court did not hold the invention patentable; it expressly left the Section 3(p) objection open, and it confined the remand so tightly that no fresh material may be filed beyond the NBA approval itself (approval that, in the event, was granted on 4 July 2024, almost five months after the refusal), and it has directed the Controller to decide within two months of receiving the order. This is a victory on process, not on merits. For anyone tempted to treat Manu Chaudhary as a green light for herbal or traditional-knowledge patents, the caution is plain: the substantive hurdles (novelty, inventive step, the Section 3 exclusions) remain exactly where they were. The applicant here prevailed, but only after a refusal, an appeal and a remand, a long and expensive route to a re-hearing she should never have needed. The lesson is not “you can win on appeal.” The lesson is that NBA and ABS compliance, and the timing of it against examination and grant, belong in the filing strategy from day one. Get that right and the question in Manu Chaudhary never arises. Get it wrong and you are litigating your way back to the starting line. That is precisely the kind of foresight good counsel is for. Would your current patent portfolio survive an NBA or Section 3 challenge? It is a question best answered before an examiner asks it. Our patents team advises on biodiversity compliance, prosecution strategy and appeals across the life sciences. Contact S.S. Rana & Co. for a considered view. a Not automatically. In Manu Chaudhary (Delhi High Court, 7 February 2026), the Court held that where an applicant has applied to the National Biodiversity Authority in good faith and placed proof of the pending application on record, the Controller should defer the decision rather than refuse the patent. But the outcome turned on the applicant having applied in time and having proved it. Whether your own file meets that standard is a fact-specific assessment best made with counsel before a hearing, not after a refusal. a The Biological Diversity Act, 2002 and the traditional-knowledge Guidelines tie NBA permission to the grant of the patent. That does not mean the timing is a free-for-all: how and when you apply to the NBA, and how you signal that to the patent office during examination, can be the difference between a deferral and a refusal. The safe answer depends on the specifics of your invention and its use of Indian biological material. Advice at the filing stage is strongly recommended. a No, but do not assume it is safe either. Section 3(p) of the Patents Act, 1970 excludes, among other things, subject matter that is in effect traditional knowledge or an aggregation of known properties. In Manu Chaudhary the Court left the Section 3(p) objection open for the Controller to decide afresh, which tells you how fact-sensitive it is. Whether your claims survive depends heavily on how the invention is framed and evidenced, a question worth putting to a patent attorney early. a Sections 128 and 132 of the Patents Act govern who may appear before the Controller, and appearance through the wrong representative can itself become an objection, as it did in Manu Chaudhary, where the Court treated it as a curable defect on remand. Relying on it being curable is a risk you need not take. Proper representation by a registered patent agent or advocate protects the application; we can advise on who should appear and when. a An appeal to the High Court under Section 117A(2) of the Patents Act may lie, as it did in Manu Chaudhary, and can result in the refusal being set aside and the application remanded for fresh consideration. But an appeal costs time and money, and whether you have viable grounds depends closely on the record and the reasons for refusal. If you are facing a refusal, speak to our patents team promptly so the appeal window is not lost.

AI in Patent Examination: Inside India’s New Guidelines from the CGPDTM

By SSR Patent Team

The Office of the Controller General of Patents, Designs and Trade Marks (CGPDTM) has issued Guidelines for the Use of Artificial Intelligence in Patent Examination on August 11, 2026. The document is notable less for what it permits than for the discipline it imposes. It treats artificial intelligence as an instrument of assistance and refuses to let it stand in for the statutory and quasi-judicial functions of the Examiner and Controller. For practitioners, the interest lies in the granularity: the Guidelines catalogue, in unusual detail, both the tasks for which AI may be deployed and the failure modes that make independent human verification indispensable. The Guidelines sit within a wider Government of India posture on responsible AI — they expressly invoke the NITI Aayog Responsible AI approach — and translate that posture into the specific setting of patent examination, a function the document rightly describes as document-intensive, technically complex, legally evolving and time-sensitive. Their stated object is to support efficiency and quality while preserving confidentiality, accountability, consistency and the independent application of mind. Their scope extends across screening, classification, search, translation support, drafting support, technical comparison and knowledge retrieval. Two structural features repay attention. First, the Guidelines distinguish public AI tools — general, consumer-facing systems trained on broad datasets — from private AI tools operating in closed environments, and record that the Patent Office’s subscribed search databases already carry in-built private AI better suited to the patent context. Second, they are candid about limitation: an entire section enumerates fourteen distinct risks, from hallucination and omission of critical detail to classification drift, black-box opacity and bias in training data. Prosecuting or defending a patent in India? The texture of examination is changing. Our patents team can advise how these Guidelines bear on a live application or a pending opposition. Speak to us before your next response falls due. The organising idea is a line between assistance and adjudication. AI is positioned to assist, and not replace, the Examiner or Controller. Any AI-assisted output that could influence search or examination must undergo manual review and validation. Critically, the officer remains fully responsible for every official act involving AI, and the Guidelines state in terms that the use of AI does not dilute or transfer that responsibility. An AI output may be adopted only after the officer is personally satisfied as to its correctness, relevance and appropriateness. For a practitioner, that principle is not merely aspirational. It is the standard against which an AI-influenced office action can be measured. If a First Examination Report rests on a classification, a prior-art mapping or a citation that the officer has not independently verified, the Guidelines supply the vocabulary — and the internal benchmark — for saying so. The Guidelines set out twelve typical uses, each paired with its risks and mandated safeguards. The pattern is consistent: AI may accelerate the preliminary work; the officer must own the conclusion. AI may suggest candidate IPC/CPC classifications, generate search vocabulary and concept clusters, and — through the Office’s subscribed private tools — assist with prior-art retrieval. The Annexure demonstrates, with worked examples, that classification output is model-, prompt- and parameter-sensitive: the same claim run through different tools produced materially different, and sometimes irrelevant, classifications. The safeguard is verification against authenticated IPC/CPC sources and a considered reading of the claims with the complete specification. Final search results, the Guidelines insist, must be selected on the officer’s independent judgment rather than accepted from the system. AI may extract claim features and offer preliminary novelty or inventive-step mapping — but this is the category the Guidelines treat with the most caution. A worked example is instructive: the same claim and the same prior art, put to the tool under differently framed prompts, yielded confident conclusions of both novelty and lack of novelty. The document’s warning is that one-to-one claim mapping demands consideration of explicit and implicit disclosures, drawings and context, and is therefore prone to significant error and omission. Any AI mapping is an assistive input only; the conclusion must rest on the officer’s own analysis. Drafting Support and Translation AI may polish grammar, structure and readability of a draft the officer has already prepared, and may give a rough understanding of foreign-language documents. The strings are precise: language support must not spill into substantive reasoning; machine translation relied upon for an objection must be checked and its use recorded in the reasoning on patentability; and every citation the tool proposes must be independently verified in the original source before use. Received an FER that doesn’t add up? Where an examination report leans on a shaky classification, an over-broad search or an unverified citation, there is a principled basis to push back. Ask our patents team to review it. AI may also assist in flagging preliminary clarity concerns in a claim — for instance, ambiguous quantitative or structural language that could attract an objection under Section 10(4) — and in surfacing possible gaps in the sufficiency of disclosure, such as missing implementation detail or an unsupported best-mode requirement. The Guidelines are explicit that both are treated as preliminary issue-spotting only: the legal determination of whether an objection is well founded, and the drafting of any resulting objection in the Office’s own style with the applicable statutory provisions, remains entirely for the officer. Among the clearest commands in the document is that unpublished patent application contents, confidential office records and internal deliberative material must not be entered into public AI tools. Claims, descriptions, prior-art notes and hearing notes are singled out. The concern is disclosure: material entered into an external system may travel beyond the approved environment, with obvious consequences for the novelty and integrity of an unexamined application. The permitted path is confined to approved internal or secure authorised environments. For applicants, this is a meaningful institutional safeguard — and one that intersects directly with wider data-protection expectations. The Guidelines list six prohibited uses. In substance: entering unpublished or confidential material into public AI tools; using AI as a substitute for the officer’s application of mind on substantive questions — novelty, inventive step, industrial applicability, sufficiency, clarity, unity; issuing office actions, FERs, hearing notices or decisions solely on AI output without adequate oversight; citing AI-suggested case law, prior art or literature without independent source verification; incorporating AI content into official communications without review, correction and adoption; and relying solely on AI for decisions affecting the rights of applicants, patentees or third parties — with opposition proceedings expressly named. Opposition on foot? The Guidelines single out opposition proceedings as a setting where AI cannot be the sole basis of decision. That has strategic weight. Our team advises on pre-grant and post-grant opposition strategy in this new environment. The administrative measures are, for the most part, enabling rather than mandatory — framed as steps the competent authority may take. They include recording the material use of AI (tool, nature of use, date), with the possibility that such records are disclosed to stakeholders to build confidence in the process; an AI Governance Committee drawn from the Examination Division, IT and QMS to approve tools, classify permitted and prohibited uses and revise the Guidelines periodically; structured training on AI’s limitations and on identifying hallucinations and fabricated citations; and provision for independent audit, impact assessment and incident reporting. An Annexure sets out an officer’s checklist and a declaration on AI use in work-product delivery. What This Means for Applicants, Patentees and Opponents For those whose rights are examined, the Guidelines are less a technical curiosity than a shift in the environment in which patents are won and defended. They create room to interrogate AI-influenced actions against a published internal standard; they protect the confidentiality of unexpected quarters of an unpublished file; and they place opposition proceedings under particular scrutiny. Whether and how to invoke any of this turns on the specifics — the stage of the matter, the technology, the reasoning on the face of the report — which is precisely where experienced counsel earns its keep. Talk to S.S. Rana & Co. From filing strategy to FER responses and opposition, our patents team can help you navigate examination in the age of AI. Get in touch to discuss your portfolio.

Mobile Phone Manufacturing Scheme (MPMS): India’s Next Leap in Electronics Manufacturing

By SSR Corporate Team

On 31 July 1995, India heard its first mobile ring, when the then Union Minister of Communications spoke from Delhi to the then Chief Minister of West Bengal in Kolkata. For most of the two decades that followed, India consumed mobile phones without making them. In 2014, the country had only two mobile phone manufacturing units and was approximately 78% import dependent, with roughly 75% of domestic demand in 2014-15 being met through imports. Today, India has over 300 mobile manufacturing units and imports account for about 0.02% of domestic demand.[1] The turnaround followed a sequenced policy design. As the Ministry of Electronics and Information Technology has explained in Parliament, the Government built the ecosystem in stages, beginning with finished products, followed by sub-assemblies, then components, and eventually machinery and tools. Make in India was launched in 2014, the Phased Manufacturing Programme followed in 2017, the National Policy on Electronics came in 2019 and the Production Linked Incentive Scheme for Large Scale Electronics Manufacturing (PLI-LSEM) was rolled out in April 2020, supported by schemes for components, semiconductors and manufacturing clusters, the Electronics Component Manufacturing Scheme approved in April 2025 with an outlay of Rs 22,919 crore, and the permission of 100% foreign direct investment in the sector.[2] Figure 1: The policy flow from India’s first mobile call to the approval of MPMS The Growth Story in Numbers The results of twelve years of this effort were placed before the Lok Sabha on 22 July 2026. Production of electronics goods rose from about Rs 1.9 lakh crore in 2014-15 to about Rs 13.11 lakh crore in 2025-26, and exports of mobile phones from about Rs 1,500 crore to about Rs 2.59 lakh crore.[3] Three structural shifts stand out.[4] India has moved from being a net importer of mobile phones in 2014 to a net exporter, 99.2% of the mobile phones used in India are now made in India, and India is the second largest manufacturer of mobile phones in the world in volume terms. Smartphones, absent from India’s top 100 exported commodities in 2014, have emerged as the nation’s single largest standalone merchandise export product, overtaking individual traditional giants like high-speed diesel and loose polished diamonds. While the United States, the United Arab Emirates, and the Netherlands serve as the primary global destinations for finished devices, China and Hong Kong have integrated tightly into this network as a multi-billion-dollar corridor for advanced smartphone components and sub-assemblies, heavily bolstered by the Electronics Component Manufacturing Scheme (ECMS).[5] The sector supports about 25 lakh direct and indirect jobs, of which the mobile ecosystem accounts for about 12 lakh, and in certain high precision segments women constitute almost 70% of the workforce. Figure 2: Growth in electronics and mobile phone production and exports On 15 July 2026, the Union Cabinet chaired by the Prime Minister approved the Mobile Phone Manufacturing Scheme with a budgetary outlay of Rs 62,500 crore. The stated objectives are to further scale up production, deepen domestic value addition, strengthen supply chain resilience and enhance global competitiveness. Importantly, the scheme also aims at building Indian brands so as to achieve technological sovereignty, capture large economic value and create Indian patents in design and research and development. The scheme takes over from PLI-LSEM, whose tenure ended on 31 March 2026.[6] On the same day, the Cabinet also approved Semicon 2.0 with an outlay of Rs 1,27,500 crore for semiconductor fabrication, advanced packaging, materials and equipment, with a special focus on design and talent.[7] Read together, the two approvals amount to a coordinated intervention across the finished product, the component and the silicon layers of the same value chain. The salient features of MPMS are set out below.[8] Monthly Sales Report Particulars Details Budgetary outlay Rs 62,500 crore Tenure 5 years, from FY 2026-27 to FY 2030-31 Base incentive 2.25% to 5% of eligible sales, at differentiated rates Domestic sourcing incentive Up to 1.5% of eligible sales, linked to domestic sourcing of key components and sub-assemblies Design and R&D incentive 3% of eligible sales, for design and research and development of the product, for building Indian brands Expected cumulative production Approximately Rs 39,00,000 crore during the scheme tenure Expected direct employment Approximately 60,000 direct jobs Table 1: Salient features of the Mobile Phone Manufacturing Scheme Continuity of policy. PLI-LSEM ended on 31 March 2026, and manufacturing decisions in this sector rest on multi year capital commitments. By approving MPMS with effect from FY 2026-27, the Government has avoided a gap between one incentive framework and the next. Reward linked to depth, not merely to volume. The additional incentive of up to 1.5% linked to domestic sourcing of key components and sub-assemblies is the provision that pushes manufacturers beyond final assembly, and it gives Indian component suppliers a demand side pull that supply side support alone could not create. A distinct incentive for Indian brands. A recurring criticism of India’s electronics success has been that the country assembles for the world without owning much of the value. The additional 3% for design and research and development, directed at building Indian brands, funds the part of the value chain where margins and long term ownership reside. Scale and predictability. A five year tenure, a defined outlay and stated rate bands allow applicants to model returns with confidence. Cumulative production during the tenure is expected to reach approximately Rs 39,00,000 crore, with around 60,000 direct jobs and a considerably larger indirect effect.[9] To our mind, the most forward looking element of the scheme is the additional 3% for design and research and development, read with the stated objective of creating Indian intellectual property covering patents, design and trademarks in design and R&D in phones.[10] This is the first time that an Indian manufacturing incentive has put money behind the creation of intangible assets rather than only behind units produced. The detailed guidelines are awaited. Our submission is that when they are notified, the design and research and development component should recognise filings of  intellectual property in India as the criterion for eligibility, and should count not only patents but also registered designs, copyright and trademarks.[11] Filings made before the Indian offices are dated, objective and verifiable, and are the clearest evidence that design and development work has genuinely been carried out in India rather than merely invoiced here. In a handset it is the design/innovation, along with the user interface and the brand, that carries the product’s identity. Such a criterion would serve both sides. It would give the administering ministry an auditable test of value addition in place of self-certified expenditure, and it would give industry a clear reason to create and retain intellectual property in India. The scheme deserves the appreciation it has received. It builds on a decade of consistent policy, it rewards depth rather than volume, and it backs the ambition of Indian brands and not Indian factories alone. Its lasting contribution may be measured not in units exported but in the Indian designs, patents and brands created during its tenure, and linking the design and R&D incentive to Indian IP filings is the surest way of securing that outcome. [1] Press Information Bureau, Government of India, Backgrounder titled “Mobiles: Catalysts of India’s Digital Rise”, dated 18 September 2025. The figure of 0.02% denotes the residual share of imports in total domestic demand. [2] Lok Sabha Unstarred Question No. 510 titled “Increase in Electronics Manufacturing”, answered on 22 July 2026 by the Minister of State for Electronics and Information Technology, Shri Jitin Prasada. The outlay of Rs 22,919 crore under the Electronics Component Manufacturing Scheme is recorded in the Press Information Bureau Backgrounder dated 18 September 2025. [3] Lok Sabha Unstarred Question No. 510, answered on 22 July 2026, Ministry of Electronics and Information Technology. Figures are approximate, as stated in the reply. [4] Lok Sabha Unstarred Question No. 510, answered on 22 July 2026, Ministry of Electronics and Information Technology. [5] https://www.aninews.in/news/business/last-year-we-exported-rs-35000-crore-components-to-china-ashwini-vaishnaw20260608185224/ [6] Press Information Bureau release titled “Cabinet approves Mobile Phone Manufacturing Scheme (MPMS)”, Release ID 2284789, dated 15 July 2026, available at https://www.pib.gov.in/PressReleasePage.aspx?PRID=2284789. The release also records that the tenure of the PLI Scheme for Large Scale Electronics Manufacturing ended on 31 March 2026. [7] Lok Sabha Unstarred Question No. 510, answered on 22 July 2026, recording the Union Cabinet’s approval of Semicon 2.0 on 15 July 2026 with a total outlay of Rs 1,27,500 crore. [8] Press Information Bureau release, Release ID 2284789, dated 15 July 2026. [9] Press Information Bureau release, Release ID 2284789, dated 15 July 2026, recording expected cumulative production of approximately Rs 39,00,000 crore and around 60,000 direct jobs during the scheme tenure. [10] Press Information Bureau release, Release ID 2284789, dated 15 July 2026, which records that the scheme aims at building Indian brands to achieve technological sovereignty, capture large economic value and create Indian patents in design and R&D. [11] The detailed operational guidelines under the Mobile Phone Manufacturing Scheme are awaited as on the date of this article. The suggestions in this section are our own and should be read as such.

CGPDTM Issues Guidelines for the Use of Artificial Intelligence in Patent Examination Procedures

By SSR Patent Team

The Office of the Controller General of Patents, Designs and Trade Marks (“CGPDTM” or “the Patent Office”) has issued the “Guidelines for the Use of Artificial Intelligence in Patent Examination Procedures” (“the Guidelines”), a comprehensive policy document regulating how Examiners and Controllers at the Indian Patent Office may use Artificial Intelligence (“AI”) and Generative Artificial Intelligence (“GenAI”) tools while discharging their statutory functions under the Patents Act, 1970. The Guidelines mark one of the most detailed attempts by an Indian statutory/quasi-judicial authority to codify the permissible and impermissible uses of AI tools in a document-intensive, technically complex, and legally sensitive administrative function. They are accompanied by two Annexures: Annexure-I, containing a non-exhaustive set of illustrative, real-world examples of AI use across classification, search, translation, drafting, novelty/inventive-step analysis, clarity objections, sufficiency-of-disclosure assessment, and case-law research; and Annexure-II, a Checklist and Declaration to be completed by officers who rely on AI-assisted output. This Alert summarises the key features of the Guidelines, the risks and safeguards they identify for each typical use-case, the categories of prohibited use, and the administrative and governance measures contemplated, together with our analysis of what this means for patent applicants, patentees, and IP right holders engaging with the Indian Patent Office. BACKGROUND Patent examination is inherently document-intensive, technically specialised, and time-sensitive, involving classification, prior-art search, claim construction, and the application of statutory tests such as novelty, inventive step, industrial applicability, and sufficiency of disclosure. Scholarly literature cited in the Guidelines recognises that AI tools can assist in suggesting search terms, retrieving and ranking relevant documents, and reducing the time and cost of screening large volumes of patent literature, and that more recent research has begun to explore AI’s role in novelty analysis specifically. The Guidelines situate this development within India’s broader policy architecture on responsible AI use in public functions, expressly referencing NITI Aayog’s “Responsible AI #AIFORALL” Approach Document for India (Part 1 and Part 2), which emphasises human oversight, accountability, safety, transparency, privacy, risk-based governance, and institutional safeguards in the deployment of AI within Government functions. The CGPDTM Guidelines translate these general principles into a sector-specific framework tailored to patent examination. Notably, the Guidelines state that the illustrative examples in Annexure-I were themselves prepared with the assistance of certain public GenAI tools and certain private tools accessible to the Patent Office, though the specific tools used have not been identified on account of their proprietary nature. Objective and Scope The stated objective of the Guidelines is to regulate the use of AI in patent examination in a manner that supports efficiency and quality while preserving confidentiality, accountability, consistency, and the independent application of mind by the Examiner or Controller. The Scope extends to AI use in screening, classification, search, translation support, drafting support, technical comparison, and knowledge retrieval — effectively covering the entire examination workflow. Key Definitions Artificial Intelligence (AI): a machine-based system capable of generating outputs such as predictions, recommendations, classifications, summaries, or content based on input data and prompts. Generative AI (GenAI): any AI system capable of generating text, images, summaries, explanations, or other content in response to prompts or instructions. Public AI tools: tools accessible to the public (free or paid consumer platforms), trained on large, diverse public datasets — versatile, but potentially limited in domain-specific depth and reliability. Private AI tools: closed/restricted systems, typically trained on proprietary or controlled datasets, with access limited to authorised personnel — considered more suitable for sensitive/proprietary data. AI-assisted output: output produced through a collaborative human-AI process with a “human in the loop” providing direction, context and oversight — as distinct from fully automated AI output. The Guidelines note that the Patent Office’s own subscribed patent search databases have in-built AI tools specifically designed for patent search, which are stated to offer better data security and better contextual understanding of patent-specific prompts compared to general-purpose public tools. Typical Uses, Risks and Safeguards Section 4 of the Guidelines sets out twelve specific use-cases of AI in patent examination, each mapped against its possible benefits, potential risks, and prescribed safeguards. The key use-cases are summarised below: Use-Case Possible Benefit Key Safeguard IPC/CPC classification Faster identification of candidate classes Must be verified from an authenticated source (e.g., WIPO-IPC/CPC) against claims read with the complete specification; adjusted manually where necessary Search terms & concept clusters Broadened search vocabulary (synonyms, related terms) Officer must select final terms with complete human oversight; AI output only a starting aid Preliminary claim-feature extraction Structured feature lists for comparison Claim must be re-read manually; no novelty/inventive-step analysis on an unverified AI feature list Translation support Rough understanding of foreign-language documents Reliance on machine translation must be expressly recorded in the reasoning for patentability Improving structure/language of draft communications Better grammar, coherence, formatting Permitted only after the officer’s substantive draft/reasoning is prepared; every statement must be checked Legal/technical citations Quick identification of case law/technical references No citation used unless independently verified against the original source; discarded if verification is not feasible Public GenAI tool + unpublished application material — Prohibited outright Preliminary novelty/inventive-step analysis Preliminary claim mapping and structured comparison Treated only as preliminary assistive input; final conclusion must rest on the officer’s own independent analysis Clarity issues in claims Preliminary issue-spotting under Section 10(4) Officer must independently determine the legal/technical basis before any objection is issued Prior-art search (officially subscribed AI tools) Improved, structured search results Queries refined manually; results assessed and selected by the officer’s independent judgment Sufficiency of disclosure Identification of possible enablement gaps Complete specification, drawings and CGK examined manually before any objection is framed Case law / legal concepts Summarisation for preliminary understanding No verbatim quote, ratio or legal proposition relied upon without checking the primary source Note: The table above is a summarised, non-exhaustive representation of Section 4 of the Guidelines. Annexure-I contains detailed worked examples — including comparative outputs from multiple AI tools for the same patent claims — illustrating model-sensitivity, prompt-sensitivity, and instances of fabricated case-law quotations, which readers engaged in patent prosecution before the Indian Patent Office may find instructive. Prohibited Uses Section 5 of the Guidelines expressly prohibits the following: Entering unpublished patent application contents, confidential office records, or internal deliberative material into public AI tools. Using AI as a substitute for the Examiner’s or Controller’s application of mind on substantive matters, including novelty, inventive step, industrial applicability, sufficiency, clarity, or unity of invention. Issuing office actions, First Examination Reports (FERs), hearing notices, decisions, or other official communications solely on the basis of AI-generated output without adequate human oversight. Citing case law, prior art, scientific literature, or other AI-suggested references without independent verification from authentic sources. Using AI-generated content in official communications without review, correction, and adoption by the concerned officer. Relying solely on AI for decisions affecting the rights of applicants/patentees or third parties, particularly in proceedings involving opposing parties (e.g., opposition proceedings). Administrative and Governance Measures The Guidelines contemplate the following institutional measures: Recording of AI use: the competent authority may prescribe recording of material AI use, including tool name, nature and date of use, for supervision, audit and quality review; such disclosure may also be made available to stakeholders. AI Governance Committee: a dedicated committee is contemplated, comprising Examiners/Controllers from the Examination Division, officers from the IT Office, and the QMS Division, nominated to cover different technology groups. Its remit includes tool review/approval, categorisation of permitted/prohibited uses, prescription of safeguards, pilot studies, review of errors/complaints/policy breaches, periodic revision of the Guidelines, and stakeholder/expert consultations. Training and capacity building: covering AI functionalities and limitations, confidentiality risks, prompt discipline, verification of AI-assisted output, and identification of hallucinations and fabricated citations. Independent audit and incident reporting: the competent authority may also provide for independent audit, impact assessment, feedback mechanisms, and incident reporting systems. Annexure-II supplements this framework with an eight-point Checklist and Declaration — a “No” answer to any checklist item means the AI output should not be relied upon without further verification, or not used at all — together with a proposed declaration module requiring officers to confirm the purpose for which AI was used (classification, search support, technical understanding, language refinement, or other permitted assistive use) in each work product. Analysis A calibrated, use-case-specific framework rather than a blanket restriction. Rather than prohibiting AI use outright, the Guidelines adopt a differentiated approach — permitting assistive use across most stages of examination (classification, search, translation, drafting support, preliminary novelty/inventive-step mapping) while confining AI strictly to a preliminary or supportive role, with the substantive statutory determination reserved exclusively to the Examiner or Controller. Confidentiality remains a central concern. The outright prohibition on entering unpublished application material into public AI tools reflects the same confidentiality anxieties that have driven comparable restrictions in other Indian regulatory contexts. Applicants and their counsel dealing with the Patent Office may wish to note that this restriction protects applicants’ own unpublished subject matter from inadvertent disclosure through Patent Office use of public AI tools, and stakeholders may consider seeking confirmation, where relevant, that only officially subscribed or approved private AI tools have been used in relation to their pending applications. Implications for prosecution strategy. The illustrative examples in Annexure-I — particularly the demonstrated model-sensitivity and prompt-sensitivity of AI-generated IPC/CPC classifications, search vocabulary, and even novelty conclusions on identical claim sets — signal that applicants and their attorneys should not assume convergence or reproducibility in AI-assisted classification or search outcomes issued by the Patent Office. This may be a relevant consideration when responding to First Examination Reports that draw upon AI-assisted search results, particularly where a cited reference appears only loosely connected to the claimed invention. Case-law citation risk cuts both ways. The Guidelines’ worked example on fabricated verbatim quotations attributed to an IPAB order is a cautionary illustration relevant not only to Patent Office officers but equally to practitioners and applicants who may themselves rely on AI tools while preparing responses, hearing submissions, or appeals. Independent verification of any AI-suggested citation against the primary source remains essential for all parties before the Patent Office, not the Office alone. Accountability is expressly non-transferable. The Guidelines are unambiguous that the use of AI does not dilute or transfer the Examiner’s or Controller’s statutory responsibility. This should reassure stakeholders that AI-assisted examination outcomes remain subject to the same standards of reasoned, individualised determination as manual examination, and that AI-generated content cannot itself be relied upon as a defence to a legally infirm office action or order. Key Takeaways for Applicants, Patentees and IP Right Holders The Guidelines confirm that AI may lawfully assist Patent Office examination across classification, search, translation, drafting support, and preliminary novelty/inventive-step mapping — but only as an assistive tool, never as a substitute for the Examiner’s or Controller’s independent application of mind. Entry of unpublished application material into public AI tools by the Patent Office is expressly prohibited; only approved internal/private tools may be used for such material. No office action, FER, hearing notice, or decision may be issued solely on the basis of unverified AI-generated output; every substantive statement, citation, and classification must be independently verified by the officer. Applicants encountering search results, classifications, or objections that appear anomalously broad, tangential, or unsupported may find it useful to specifically probe whether — and how — AI tools were used, given the documented model- and prompt-sensitivity of such outputs. The contemplated AI Governance Committee, recording/disclosure obligations, and Annexure-II declaration module indicate that greater transparency around AI use in individual prosecution files may become available to stakeholders over time. Practitioners should treat the Guidelines’ cautionary findings on hallucinated case-law citations as equally applicable to their own use of AI tools in drafting submissions, responses, and appeals before the Patent Office and appellate fora. The Guidelines represent a significant and welcome step toward transparent, risk-calibrated integration of AI into India’s patent examination system, and are likely to inform similar frameworks across other IP and regulatory functions in India. We are tracking further developments, including the constitution of the AI Governance Committee and any subsequent revisions to the Guidelines, and will update this Alert as fresh information becomes available.

DPIIT Liberalises FDI Policy to Permit Inventory-Based E-Commerce Exclusively for Exports: The Foundational Change Behind DGFT’s New Framework

By SSR Corporate Team

Press Note No. 3 (2026 Series), dated July 23, 2026, Department for Promotion of Industry and Internal Trade, FDI Policy Cell The Department for Promotion of Industry and Internal Trade (“DPIIT”), Ministry of Commerce & Industry, has issued Press Note No. 3 (2026 Series) dated July 23, 2026 (the “Press Note”)[1] , reviewing the Foreign Direct Investment (“FDI”) policy on the e-commerce sector to permit the inventory-based model of e-commerce, on a restricted basis, for exports of domestically manufactured and/or produced goods. The Press Note was issued by Dr. Jai Prakash Shivahare, Joint Secretary to the Government of India, under DPIIT F. No. P-15015/9/2025-FDI Policy dated July 23, 2026. This Press Note is the foundational policy change underpinning the DGFT’s subsequent rollout of the Inventory-based Cross-border E-Commerce Facilitation Framework — namely, Notification No. 27/2026-27[2] and Public Notice No. 25/2026-27[3] , both dated August 5, 2026, covered in Parts II and I respectively of this series. Notification No. 27/2026-27 expressly cross-refers to “Para 5.2.15.2.5 of Consolidated FDI Policy as amended vide Press Note No. 3 (2026 Series) dated 23.07.2026” as the source of the FDI liberalisation that enables e-commerce entities to undertake export-only inventory operations under the DGFT Framework. BACKGROUND Under the Consolidated FDI Policy Circular of 2020[4] dated October 15, 2020, as amended from time to time (the “FDI Policy”), FDI is permitted in the Business-to-Business (B2B) and marketplace models of e-commerce, but is not permitted in the Business-to-Consumer (B2C) e-commerce model or the inventory-based model of e-commerce, where inventory of goods and services is owned by the e-commerce entity and sold directly to consumers. With a view to facilitating greater exports through easier and increased access to global markets by Indian sellers, DPIIT has reviewed the FDI Policy and decided that the restrictions on the inventory-based model of e-commerce shall not apply in the case of exports of domestically manufactured and/or produced goods/products. New Para 5.2.15.2.5 — Inventory-Based Model of E-Commerce Exclusively for Exports A new paragraph is inserted under Para 5.2.15.2 of the FDI Policy, reading as follows: An e-commerce entity is permitted to engage in the inventory-based model of e-commerce exclusively for the export of goods/products manufactured and/or produced in India, as per the applicable provisions of the Foreign Trade Policy 2023, read with the Handbook of Procedures (HBP), and the Foreign Exchange Management (Export of Goods & Services) Regulations, 2015, as amended from time to time. The restrictions on the B2C and inventory-based model of e-commerce stipulated under Para 5.2.15.2.1 to Para 5.2.15.2.4 of the FDI Policy shall not apply to the export of goods/products through e-commerce as permitted under the new Para 5.2.15.2.5(i). Effective Date The decision is stated to take effect from the date of the corresponding FEMA notification.[5] Consequential Action Directed by the Press Note The Joint Secretary, Department of Economic Affairs, has been asked to suitably incorporate the policy changes in the Foreign Exchange Management (Non-Debt Instruments) Rules, 2019 , the relevant schedules thereof, and the FIRMS portal. The Reserve Bank of India, Foreign Exchange Department, has similarly been asked to incorporate the policy changes in the Foreign Exchange Management (Non-Debt Instruments) Rules, 2019[6], the relevant schedules thereof, and the FIRMS portal. The Press Information Officer has been asked to give wide publicity to the Press Note, and the NIC Section, DPIIT, has been asked to upload it on DPIIT’s website. Significance for E-Commerce Entities with Foreign Investment This Press Note is the source of the FDI liberalisation without which the DGFT’s Inventory-based Cross-border E-Commerce Facilitation Framework could not extend to e-commerce entities with foreign investment. Prior to this amendment, an e-commerce entity with FDI could not itself own or hold inventory of goods for sale — whether to domestic consumers (B2C) or, on a plain reading of the erstwhile restriction, potentially even for export purposes structured as an inventory model. The new Para 5.2.15.2.5 carves out a narrow, export-only exception: an e-commerce entity with FDI may now hold inventory, but only of goods manufactured or produced in India, and only for export. Critically, as Notification No. 27/2026-27 subsequently clarified, an e-commerce entity availing of this carve-out must undertake such operations through a separate legal entity incorporated for the purpose, and must disclose its shareholding pattern and the nature of its ownership or control relationship with the e-commerce entity at the time of registration (or amendment) as an Exporter-on-Record. Entities structuring such operations should treat the FDI Policy carve-out and the DGFT registration/disclosure requirements as an integrated compliance exercise rather than two separate approvals. Timing and Sequencing Clients should note the sequencing across the three instruments in this series: the FDI Policy carve-out (this Press Note, dated July 23, 2026) came first, followed roughly two weeks later by the DGFT’s substantive FTP amendment (Notification No. 27/2026-27) and its accompanying operational Public Notice (No. 25/2026-27), both dated August 5, 2026. Since the Press Note ties its own effective date to a further FEMA notification, entities should confirm that the underlying FEMA amendment (to the Foreign Exchange Management (Non-Debt Instruments) Rules, 2019) has in fact been issued before treating the FDI carve-out as operative, notwithstanding that the DGFT Framework itself has already been notified with immediate effect. Press Note No. 3 (2026 Series) is the policy trigger for India’s new inventory-based cross-border e-commerce export regime, opening a narrow, export-only exception to the long-standing FDI restrictions on inventory-based e-commerce. Read together with Notification No. 27/2026-27 and Public Notice No. 25/2026-27 (Parts II and I of this series), it completes the regulatory picture: the FDI Policy basis for foreign-invested e-commerce entities to participate, the FTP framework governing Exporters-on-Record and Sellers-on-Record, and the operational procedures for registration and compliance. E-commerce entities with foreign investment considering export-only inventory operations are advised to verify the status of the corresponding FEMA notification and to structure their proposed operations, corporate arrangements, and disclosures consistently across all three instruments. For further information or assistance in relation to the FDI Policy carve-out for inventory-based e-commerce exports or the related DGFT Framework, please reach out to S.S. Rana & Co. [1] https://www.dpiit.gov.in/static/uploads/2026/07/ceb0cae74fd4e83094dc6b50c3d53f92.pdf [2] https://content.dgft.gov.in/Website/dgftprod/c6898e09-daac-4df8-a981-d3124f047547/Notif%2027%20-%20English.pdf [3] https://content.dgft.gov.in/Website/dgftprod/a0a2dcc6-b0f7-4b4d-8163-abbe3e404689/PN%2025-%20English.pdf [4] https://www.dpiit.gov.in/static/uploads/2025/07/6457fc2703ee6082366c4a958b6473a8.pdf [5] FEMA notification is not publicly available, please verify its content before relying on the data. [6] https://www.enforcementdirectorate.gov.in/media/fema/ab93a739-71e4-4db5-977d-d6652850de2d_Foreign%20Exchange%20Management%20(Non-Debt%20Instrument)%20Rules,%202019%20-%20without%20amendment_2.pdf