Indian information technology companies, and the integrators among them, stand to gain from enterprise adoption of artificial intelligence without carrying the capital burden of building computing capacity, according to a research report by brokerage Anand Rathi published on 28 October. The report argues that Indian IT offers an asset-light route into the next phase of AI spending, as businesses move from testing AI tools to putting them to work.
Indian IT offers relatively asset-light exposure to the subsequent enterprise deployment cycle, supported by net-cash balance sheets, low capex intensity and strong cash conversion,” the report said. In plain terms, Indian IT firms hold cash rather than debt, spend little on physical assets relative to revenue, and turn profit into cash at a high rate. That balance sheet shape lets them sell AI deployment work without funding the data centres and chips underneath.
The distinction matters because the costs of AI infrastructure have grown sharply. Companies building and operating AI computing capacity face large funding requirements, long commitments for computing power and the risk that some capacity goes unused. Anand Rathi says that capital intensity, funding requirement and utilisation risk remain concentrated in the infrastructure layer, while the deployment work sits elsewhere.
Why integrators are in demand
Enterprise AI adoption is moving beyond experimentation toward practical implementation, and that shift creates demand for systems integrators. Businesses adopting AI need help connecting models with existing software, redesigning workflows and putting models into daily use. Anand Rathi reads that demand as an opening for model-agnostic integrators, which help enterprises select, integrate and deploy AI models without tying the customer to one model provider.
The brokerage points to Anthropic’s spending as evidence of where that demand comes from. Anthropic’s sales, marketing and partnership expenses rose roughly 27 times year on year to $3.2 billion in the first half of calendar year 2026, driven largely by a $2.7 billion increase in channel-partner fees. Channel partners carry a product into customer organisations, and a jump of that size suggests model makers are leaning on partners to reach enterprises.
Anthropic’s customer numbers point the same way. The company had about 11,900 customers generating more than $100,000 in annual revenue as of September 2026, and around 80 percent of its first-half revenue was consumption-based, meaning customers paid in line with how much they used. Selling to that many customers, and supporting their varied systems, is work integrators have done for decades in earlier waves of enterprise software.
What the integrators argument rests on
Anthropic’s own results show both sides of the infrastructure story. The company reported revenue of $16.3 billion in the first half of calendar year 2026, a 14-fold year on year increase, while adjusted operating profit turned positive at $1.6 billion. Those figures show a model business reaching large scale and making money on operations. At the same time, Anthropic had approximately $321 billion in leases yet to commence, future commitments tied to computing capacity.
Anand Rathi uses that contrast to frame its case for Indian IT. The revenue growth shows enterprises are buying AI in earnest, while the lease commitments show how much capital the infrastructure layer must carry to serve them. A firm that helps enterprises deploy AI, without signing those leases, keeps its balance sheet light while still sharing in the spending that enterprise adoption creates.
The report also notes that Indian IT firms, acting as integrators, can help enterprises deploy AI solutions without replicating the underlying infrastructure investments. That is the core of the asset-light claim: the customer gets a working AI system, the model maker supplies the model, the infrastructure provider supplies the computing power, and integrators connect the three inside the customer’s business.
Risks integrators still face
The report carries a warning alongside its case. AI-led pricing deflation could cut the other way, Pivot News noted in the research, since AI tools can reduce the amount of human effort a customer needs and put pressure on the rates integrators charge. If customers expect AI to do more of the work for less money, the same technology that creates deployment demand could also squeeze the price of that work.
Utilisation risk is another open question. Infrastructure providers carry it today, but integrators carry a version of it too: staff hired for AI projects must stay busy as projects end and new ones begin.
For now, Anand Rathi’s reading is that the balance favours Indian IT, with asset-light exposure to enterprise AI deployment, net-cash balance sheets and strong cash conversion behind it. The $321 billion in leases yet to commence at Anthropic shows how much capital the other side must carry.