The Andhra Pradesh cabinet cleared a power distribution licence route for Google's Visakhapatnam data centre campus in April 2026, the state's first such authorisation to a private non-utility operator. The instrument sits under Section 14 of the Electricity Act 2003, has been available to every state since 2003, and was first deployed for a data centre in Uttar Pradesh in 2023; Andhra Pradesh extends it to hyperscale scale. What did the cabinet actually clear, what can other states replicate, and what in this is genuinely new rather than a precedent already set?
What was cleared in April 2026 was a state cabinet authorisation enabling Raiden Infotech India Private Limited, the implementing subsidiary, to operate as a distribution licensee for the Visakhapatnam data centre campus and the associated developments at Adavivaram, Tarluvada, and Rambilli. The enabling step operates on a specific statutory limb: under the Distribution of Electricity Licence (Additional Requirements of Capital Adequacy, Creditworthiness and Code of Conduct) Rules 2005, as amended in 2022, the minimum area of supply is a municipal corporation, three adjoining revenue districts, or a smaller area as the appropriate government may notify, and the state notifies the campus as that smaller area, so that an application confined to the data centre premises can be entertained at all. The cabinet's approval is the political-policy gate; the operative grant of licence under Section 14 of the Electricity Act 2003 is made by the Andhra Pradesh Electricity Regulatory Commission, which conducts the procedural process of public notice, examination of objections, and determination of conditions under Sections 15 and 16. These two steps, the cabinet decision and the regulatory grant, are commonly conflated in public reporting; the distinction matters because the cabinet's role is to set policy and signal political backing, while the SERC's role is to determine the conditions of licence that govern operations.
The Electricity Act 2003 placed distribution on the same architectural footing as transmission and trading: a licensed activity, under Section 14, granted by the State Electricity Regulatory Commission. The Constitution places electricity in the Concurrent List at Entry 38 of List III, which means both Centre and states can legislate; the central law is the Electricity Act 2003, but the state controls administration through the SERC and policy direction through the powers reserved under Section 108 of the Act. The instrument that Andhra Pradesh has used has therefore been available to every state for over two decades.
The existing private discoms in India, BSES Rajdhani and BSES Yamuna in Delhi, Tata Power Delhi Distribution, Adani Electricity Mumbai, CESC in Kolkata, Torrent Power in Ahmedabad and Surat, are all electricity utilities. They received their licences either through pre-Independence private operations carried forward under the first proviso to Section 14, or through state-led privatisation of unbundled state electricity boards. The first fresh distribution licence to a non-utility operator did not come in Andhra Pradesh: in June 2023 the Uttar Pradesh Electricity Regulatory Commission granted a parallel distribution licence to NIDP Developers Private Limited for its data centre park at Greater Noida, after the Uttar Pradesh government notified the smaller-area exemption that the licensing rules require. What the Andhra Pradesh matter changes is not the category but its scale and entrant: a hyperscale campus at gigawatt load, anchored to a flagship foreign-investment commitment, rather than a single developer's data-centre park. The licence does not displace Andhra Pradesh Eastern Power Distribution Company Limited as the area discom; it carves out a defined area of supply within which the operator's subsidiary will function as the licensed distributor. Within that area, the operator can directly procure power from generators, integrate renewable supply on terms it sets, build and operate the distribution network to the data halls or take open access on the state transmission system as the regulator approves, and manage reliability to the standard the operation requires. Outside that area, APEPDCL continues as before. The architecture is parallel licensee, not displacement.
The reason no state had used this instrument for a hyperscale industrial operator before is not statutory. It is fiscal. State distribution companies in India remain financially stressed; the Revamped Distribution Sector Scheme is the central instrument designed to address the Aggregate Technical and Commercial losses that drive this. Discom finances depend on industrial-tariff cross-subsidies to subsidised consumer categories, principally agriculture and below-poverty-line domestic households. A 1 GW industrial customer is not a marginal account; at hyperscale data-centre load factors, a campus of this size generates revenue that funds a meaningful share of a state discom's industrial cross-subsidy pool. Releasing such a customer from the discom's billing relationship is therefore a fiscal decision the state takes only when the commercial terms of the inbound investment justify the foregone cross-subsidy.
Andhra Pradesh's calculus is specific. The Visakhapatnam campus is a USD 15 billion commitment over five years, among the largest single foreign direct investment commitments recorded in India, with construction commencing on 28 April 2026 and a planned 1 GW build-out across three locations. The state has chosen to forgo the discom-tariff revenue that the campus would have generated as a conventional industrial consumer, in exchange for the FDI commitment and the ancillary investment the campus is expected to anchor. This is a regulatory carve-out, not a fiscal incentive; the architecture works by removing the campus from one regulatory regime and placing it inside another, rather than by paying capital subsidy out of the consolidated fund. The fiscal cost is foregone cross-subsidy revenue, absorbed by the state discom and not expressed as a budget-line expenditure.
The replicability question turns on this calculus, not on the statute. Every state can issue a Section 14 licence; few states can absorb the cross-subsidy loss that a hyperscale carve-out implies. States with healthier discom positions sit in a different room than states where agricultural cross-subsidy is the binding constraint, and any carve-out from that obligation requires either a fresh tariff settlement or a state government commitment to fund the discom directly. The instrument is statutorily uniform across India; the conditions under which a state can deploy it are not. Andhra Pradesh's discom finances, while not strong by absolute standards, sit at a position where the trade is bearable; the calculation in another state with a different agricultural-tariff structure would not produce the same answer.
What the architecture allows other states to replicate is the procedural mechanism: cabinet authorisation followed by SERC grant under Section 14, with an area of supply carved out within an existing utility's territory. What it does not allow them to replicate, without independent fiscal action, is the foregone cross-subsidy absorption. The Singapore-style or Texas-style retail-choice model, where industrial customers freely select their distributor, sits on a different regulatory architecture from what India operates; the Indian model retains the cross-subsidy obligation as the state discom's responsibility, and any carve-out from that obligation is a state-level fiscal decision, not a regulatory simplification.
For the inter-state competition that has dominated industrial allocation since the Production Linked Incentive era, the implication is that fiscal-incentive bidding has reached a structural limit. Capital subsidies, stamp duty waivers, electricity duty exemptions, and SGST reimbursements are scheme-level instruments that have converged across states; the differentiation has compressed. Regulatory carve-outs of the kind Andhra Pradesh has authorised are a different instrument class, available only to states willing to absorb the discom-finance consequences and politically able to defend the trade-off. The states that hold this instrument credibly are the states that will compete for the second wave of hyperscale infrastructure, where the binding constraint is no longer capital subsidy but power architecture, water availability, and the speed at which the state can resolve industrial-scale utility procurement on terms the operator sets.
The Visakhapatnam matter is therefore not a one-off accommodation; it is the opening move in a competition for hyperscale infrastructure that other states will now have to decide whether they can match. The Section 14 instrument has always been there. The question is which state's discom finances permit it to be used next.