One of the key aspects that has often come up in recent conversations is India’s relatively smaller share in the overall AI push seen globally. However, given the recent spurt in investment for AI datacentres, can there be additional sops expected in the upcoming Budget? Industry and economic observers believe that the upcoming Union Budget is being seen as a make-or-break moment for the country’s AI ecosystem.
Industry expectations, outlined in the Deloitte Budget Expectations 2026 report, converge around a single idea: without domestic infrastructure, sustainable scaling and indigenous models, India’s AI momentum risks remaining dependent on foreign systems.
Budget 2026 AI wishlist: Tax holiday for data centres?
At the heart of the wishlist is a push to dramatically expand domestic AI infrastructure, particularly data centres and computing capacity. The report highlights the need for India to host personal and AI-training data within its borders, in line with the Digital Personal Data Protection (DPDP) framework.
To make this viable, stakeholders are seeking long-term fiscal support, ranging from conditional tax holidays for data centre developers to full GST input tax credit or refunds on capital-intensive assets such as construction, power and electrical systems.
The report further calls for a multi-year customs duty waiver on imported critical equipment, which includes GPUs, UPS systems and cooling infrastructure. These equipments currently pushes up the project costs.
The report added that a compute-credit scheme would essentially allow the Indian startups and research labs to access discounted GPU or TPU hours on domestic infrastructure. This will reduce the dependence on offshore cloud capacity and lower barriers to model training.
Budget 2026 AI wishlist: Incentives for adoption
As AI workloads surge, the capacity expansion without sustainability guardrails could lock India into a high-carbon path, the Deloitte report warned. It further added that incentives that explicitly reward efficiency and clean energy adoption would counter this issue.
One expectation is to link accelerated depreciation and GST benefits to verifiable benchmarks such as a power usage effectiveness (PUE) of 1.4 or lower, alongside time-bound commitments to source at least half of a data centre’s power from renewable energy.
Interest subventions routed through infrastructure-focused lending institutions are seen as critical to nudging developers towards renewable power and water-neutral cooling technologies, without making projects financially unviable.
Deloitte on indigenous models and talent pipeline
Apart from hardware, India’s AI future depends on building culturally intelligent and India-first large language models, the Deloitte report added. This will require incentives for startups to procure hardware locally and train the models on domestic datasets. Such an approach, it argues, would reduce reliance on foreign foundational models and improve relevance for Indian users.
Human capital is the other missing link. The report argued that a project-based AI learning should be embedded across the education system to tackle theoretical knowledge and industry-ready skills.
On the hardware front, it also recommends relaxing procurement norms that currently favour large players, such as minimum GPU requirements per bidder, so that MSMEs and smaller firms can scale up computing capacity in phases, supported by import-duty benefits.
Big tech bets billions on India’s AI backbone
Apart from the government’s push on AI, big tech companies are increasingly acknowledging the AI potential that India holds. Microsoft led the pack with a cumulative $20.5 billion commitment, including a fresh $17.5 billion announcement in December to expand hyperscale cloud and AI infrastructure aligned with India’s data sovereignty goals, building on a $3 billion pledge earlier in the year, according to company disclosures.
Google followed with a $15 billion investment to set up an AI-focused data hub in Visakhapatnam, its largest data centre investment outside the US, positioned as a regional backbone for generative AI services across Asia-Pacific, as reported by multiple policy and industry trackers.
Amazon Web Services committed $12.7 billion in cloud and AI-ready infrastructure investments in India through 2030, while Reliance Industries began building what it describes as a 1-gigawatt AI data centre complex in Jamnagar, with plans to scale up to 3GW and an estimated $20–30 billion investment by 2027 to power its JioBrain AI platform.
What Budget 2025–2026 accounted for
The expectations around AI found partial reflection in last year’s budget, which significantly stepped up public spending on the sector. Guided by the vision of “Making AI in India and Making AI Work for India, the Cabinet had approved the IndiaAI Mission in March 2024 with an outlay of Rs 10,371.92 crore spread over five years, positioning it as the government’s flagship vehicle for building domestic AI capabilities, according to data from the Press Information Bureau (PIB).
Since its launch, the mission has moved quickly on the infrastructure front. Against an initial target of 10,000 GPUs, India has already scaled up to 38,000 GPUs, aimed at providing startups, researchers and academic institutions access to affordable, high-performance computing resources, PIB data shows.
In the Union Budget 2025-26, the IndiaAI Mission received the single largest boost, with Rs 2,000 crore earmarked for FY26.
Centre of Excellence (CoE) in AI- A reality check
The budget also announced a fresh Centre of Excellence (CoE) in AI for the education sector, with a capital outlay of Rs 500 crore, in addition to three existing CoEs. These centres are meant to drive interdisciplinary research and explore the use of AI in governance and public service delivery. Separately, allocations for the existing AI Centres of Excellence were raised to Rs 200 crore, up 82% from the revised estimate of the previous year.
Taken together, the finance ministry allocated Rs 4,349.75 crore in FY26 to schemes that directly or indirectly involve AI, including broader programmes such as the National Mission on Interdisciplinary Cyber-Physical Systems and R&D initiatives in electronics and information technology.
The wishlist for AI ahead of Budget
Taken together, the expectation and wish list ahead of Budget 2026 signal continuity rather than a sharp policy turn on AI. While allocations under the IndiaAI Mission and related programmes point to a steady expansion of research capacity, industry is still awaiting clearer fiscal levers on data centres, sustainability-linked incentives and indigenous model development.
Whether these gaps are addressed in the full budget fine print will determine how quickly India’s AI ambitions translate into on-ground capacity creation.