Digital infrastructure company Lightstorm (the company) on Tuesday announced the launch of the AI Superhighway and said it had secured a Rs 25 billion (bn) debt facility led by IndusInd Bank to fund the project. The company described the network as a utility-grade, deep trenched, high-capacity fibre system that will link Indian AI hubs with hyperscale centres in Singapore and Malaysia via the I-2SEA subsea cable system. The financing carries a 10-year tenor and is fully underwritten. Lightstorm said the initiative responds to the emergence of India as a major AI compute market, noting that hyperscalers are deploying data centres and graphics processing unit clusters across the country at an accelerating pace. The company added that infrastructure connecting that compute must match the ambition of such deployments. Domestically the network will span the southern corridor to interconnect designated AI zones, while the international leg will create a direct route to Singapore and Malaysia. The firm's SmartNet AI Fabric will provide purpose-built, low-jitter, loss-optimised transport across more than 30,000 km of terrestrial fibre and more than 21,000 km of subsea fibre, with connectivity into over 100 data centres globally. The offering is delivered through the Polarin network-as-a-service platform, enabling enterprises, hyperscalers and AI infrastructure providers to provision and monitor global connectivity on demand. The platform is designed to provide visibility into network health and performance on a real time basis. Company executives framed the project as a strategic step to support the next phase of growth in AI infrastructure and said the funding will enable expansion of both terrestrial and subsea capacity. The AI Superhighway is intended to create a high-capacity route for compute workloads and to improve transport characteristics required for large scale machine learning operations. Lightstorm expects the network to be used by cloud providers, data centre operators and AI infrastructure vendors as they scale in the region.