Five years of category-building, two years of Department for Promotion of Industry and Internal Trade (DPIIT) consultation, eleven million square feet of industrial leasing in a single quarter from Blinkit, Zepto, and Instamart, and not one ministry has issued a settled classification. The facility holds inventory, accepts orders through a digital platform, and dispatches groceries to consumers within fifteen minutes; no customer crosses the threshold. Each regulatory framework reads the same facility differently; the operator absorbs the difference. What does this reveal about how India's regulatory architecture handles asset classes it was not designed to recognise?
Dark stores are, by operational definition, 3,000 to 5,000 square foot enclosed facilities in dense urban catchments that hold inventory, accept orders through a digital platform, and dispatch last-mile delivery within a 10 to 15 minute window. No customer enters the premises. No retail transaction occurs within the four walls. The inventory moves from a warehouse-style storage configuration directly into consumer hands through a rider. By every physical and operational test, the facility is neither a warehouse nor a retail outlet. It sits in the space between, and it does not match any category the Indian regulatory architecture has previously been designed to classify.
Multiple regulatory frameworks apply to the same facility, and each arrives at a different classification.
Shops and Establishments registration is administered at the state level. Most states register a dark store as a "commercial establishment" under statutes designed for retail shops and offices, which triggers working-hours, leave, and wage provisions drafted for customer-facing retail and clerical workplaces. The facility, however, operates closer to a warehouse in function, with continuous night operations and a workforce composition unlike either archetype. The operator's compliance cost varies by state, and the same facility can hold one registration under one state's reading and a different registration under another's.
Food safety classification under the Food Safety and Standards Act 2006 governs every dark store that handles grocery, fresh produce, dairy, or packaged food, which in practice is every dark store. The category system of the Food Safety and Standards Authority of India (FSSAI), built for manufacturers, retailers, restaurants, and storage operators, does not fit the dark store configuration. A facility that stores food briefly before onward dispatch is not a conventional storage operator; it is not a retailer because no customer crosses the threshold; it is not a restaurant because no food is prepared; it is not a manufacturer. Depending on turnover, the operator registers under the central FSSAI licensing regime or the state FSSAI authority, and the licensing category applied varies between the two. The food safety compliance architecture of the dark store, the category under which it is licensed, the labelling standards that apply, the storage temperature requirements, the shelf-life tracking obligations, varies by licensing category, and the licensing category varies by which state authority is reading the same facility.
Goods and Services Tax (GST) treatment depends on whether the facility is a "place of business" for the retail sale to the end consumer or a "place of supply" for the inventory movement between the central warehouse and the dark store. The distinction determines input tax credit, registration obligations in each state of operation, and the treatment of stock transfers. The Central Board of Indirect Taxes and Customs has not issued a circular addressing the dark store configuration; state-level assessing officers apply their own interpretations, and the same operator can face different GST treatment of identical transactions across different states. The question also intersects with Section 52 of the Central Goods and Services Tax Act, which requires "e-commerce operators" to collect Tax Collected at Source on supplies made through their platforms. Whether the quick commerce platform is an e-commerce operator facilitating supply by the dark store or is itself the supplier operating through the dark store is a distinction that the statute does not resolve cleanly for the 10 to 15 minute delivery model. A platform operating dark stores across multiple states encounters the Input Service Distributor framework, cross-state Input Tax Credit utilisation, and TCS applicability as three open questions on the same set of transactions, and the answers are being composed by state-level assessing officers without a unifying CBIC position.
Foreign direct investment in logistics and warehousing is permitted up to 100 per cent under the automatic route; foreign direct investment in multi-brand retail is permitted up to 51 per cent under the government route, with state concurrence, sourcing conditions, and investment thresholds. A dark store operated by a foreign-owned quick commerce platform arguably sits in both categories and cleanly in neither. The operators have proceeded on the logistics classification. The reclassification risk sits permanently open, and the first adverse determination by any regulatory authority would cascade across every foreign-backed quick commerce platform operating in the country.
Press Note 3 (2020 Series, as amended by Press Note 2 of 2026) imposes a government-route requirement on investment from entities in countries sharing a land border with India or whose beneficial ownership trails to such countries. For quick commerce platforms whose cap tables include investors from entities touching land-border beneficial ownership, Press Note 3 compliance already applies to the investment itself. The dark store classification question compounds this. If the dark store is classified as logistics, the underlying FDI treatment is governed by the logistics sectoral framework with Press Note 3 as the overlay. If the dark store is reclassified as multi-brand retail, two layers of friction attach to the same capital: the multi-brand retail sectoral constraints, and the Press Note 3 overlay on the land-border beneficial ownership. The dark store classification question is not merely a sectoral classification question; it is the hinge on which the FDI architecture of the entire foreign-backed quick commerce sector hangs, and its resolution has consequences that no individual operator has the institutional ability to shape.
Platform worker classification under the Code on Social Security 2020 introduces a distinct regulatory architecture for the rider workforce that is the dark store's substantive labour footprint. The Code defines "platform work" and "gig worker" as distinct categories, separate from the conventional employment relationship the Factories Act or the Shops and Establishments Acts were drafted around. State-level implementation is proceeding unevenly: Rajasthan enacted the first state framework in 2023; Karnataka followed in 2025 with the Platform-Based Gig Workers (Social Security and Welfare) Act and notified rules in November 2025, levying a welfare fee of one to five percent of each transaction payout to gig workers, mandating a state Welfare Board, registration obligations on aggregators, and a two-tier grievance redressal system; Bihar enacted its own statute in 2025; Jharkhand became the fourth state with a similar law in August 2025; Telangana has a draft bill in consultation. The compliance architecture is therefore simultaneously central (the Code on Social Security itself) and state-specific (each state's platform worker legislation), and the operator must hold every state framework in view while the specific instruments of each state are still being notified. The rider workforce is regulated under a framework that did not exist when the dark store configuration first entered Indian commerce, is still being composed across central and state levels, and is likely to be the layer of fastest regulatory evolution over the next several years.
Municipal zoning was written for two commercial archetypes, the warehouse park on the urban periphery and the retail high street at the neighbourhood centre. A 4,000 square foot facility with continuous rider traffic, loading activity through the day and night, and inventory movement in a residential catchment does not match either archetype. Municipal corporations have responded variously, with some treating dark stores as warehouses requiring industrial or mixed-use zones, others as commercial establishments permitted in retail-zoned areas, and others leaving the classification to local development control officer discretion. The inconsistency is not the result of deliberate divergence; it is the result of each municipal authority applying a bye-law architecture that was written before the asset class existed.
The Department of Economic Affairs' harmonised master list of infrastructure sub-sectors determines long-term financing cost. Infrastructure classification unlocks priority sector lending treatment, specific income tax deductions, external commercial borrowing access on favourable terms, and eligibility for infrastructure investment trusts. Warehousing was added to the list in 2012. Data centres were added after sustained industry representation. Dark stores are not on the list. The capital deployed in them is treated as ordinary commercial real estate for lending and tax purposes, and the operator's cost of capital is higher than it would be if the asset class were recognised.
DPIIT has convened consultations on quick commerce classification across multiple rounds since 2024, engaging the Ministry of Electronics and Information Technology (MeitY), CBIC, DEA, state commerce departments, and, in later rounds, FSSAI and the Ministry of Labour and Employment. No ministry has taken ownership of issuing the dedicated rules. The consultation has yielded institutional awareness of the problem without creating institutional responsibility for solving it. Each ministry participating in the consultation reads the dark store through its own statutory framework, confirms that the framework does not fit cleanly, and leaves the meeting. No Press Note has been drafted. No inter-ministerial harmonisation note has been circulated. No Cabinet decision has been initiated. The consultation register remains open; the notification register remains empty.
This is the institutional pattern, not a one-off delay. The same pattern is visible in data centres, commercial space, and drones. In each, policy intent preceded classification update. In each, the classification exercise ran backward from the commercial reality rather than ahead of it. In each, the catch-up happened asymmetrically, with DPIIT recognising one dimension, DEA another, and the sectoral regulator a third, over a period of years. The dark store case is sharper because the commercial scale is already substantial. Colliers India's Q1 2026 data records 11 million square feet of industrial and warehousing leasing across the top eight cities in a single quarter, a 22 per cent year-on-year increase, with e-commerce and third-party logistics players driving the uptake; the dark store and micro-fulfilment networks of Blinkit, Zepto, and Instamart are scaling rapidly to meet 10 to 15 minute delivery timelines. The capital is deployed. The asset class is at scale. The regulatory architecture has not yet yielded a single unified classification, and the operator is absorbing the ambiguity into cost of capital, compliance expenditure, FDI structuring, and the institutional cost of simultaneously answering to multiple frameworks that each read the same facility differently.
The Indian regulatory architecture is accretive, not designed. It was not built to recognise new asset classes. It was built to process known categories. When a new category arrives, the recognition exercise is triggered by friction, not by surveillance, and the friction is borne by the operator in proportion to the capital it has already deployed.