A £202 million state-backed guarantee has helped unlock roughly £300 million of lender finance for DataVita, giving the Scottish data-centre operator capital to expand its existing Lanarkshire site and build a second facility within the region’s AI Growth Zone.

The financing has been provided by a syndicate including ING, ABN AMRO, Santander, the Scottish National Investment Bank, and Siemens Financial Services, with the National Wealth Fund guaranteeing £202 million of the debt. The structure therefore combines public balance-sheet support with private lending rather than representing a £300 million government grant.

DataVita says the money will support the expansion of DV1 and construction of a second North Lanarkshire data centre, with the capacity of both sites contracted to CoreWeave under a 15-year agreement. That long-term customer commitment is central to the financing because data-centre debt depends heavily on predictable utilisation and counterparty strength.

The deal moves the Lanarkshire AI Growth Zone from policy ambition towards physical infrastructure, although it only solves one part of the build. Data centres of this scale still depend on planning, electrical connections, cooling, equipment delivery, and the availability of enough power to keep dense AI hardware operating continuously.

Public finance moves into compute

The National Wealth Fund’s involvement reflects a wider shift in UK industrial policy, where public finance is increasingly being used to crowd private capital into infrastructure considered strategically important. AI compute has joined transport, energy, and manufacturing as an area in which ministers are prepared to use guarantees and other financial tools to accelerate investment.

For lenders, the guarantee reduces part of the credit risk attached to a capital-intensive project whose revenue arrives over many years. For DataVita, it improves access to financing at a point when the cost of power infrastructure, cooling systems, construction, and specialised data-centre equipment can make new capacity difficult to fund from company resources alone.

The CoreWeave contract also gives the project an unusually direct link to current AI demand. Rather than building speculative capacity and searching for tenants later, DataVita has a large compute customer attached to both sites, which strengthens the commercial case while also concentrating exposure to one fast-growing market.

Capital does not solve the power question

Even with financing secured, the most difficult constraint may still sit outside the balance sheet. Large AI data centres require substantial and dependable electricity supply, and Scotland’s infrastructure ambitions have already run into questions over grid connections, renewable generation, and whether local energy plans can be delivered on the timescales attached to growth-zone policy.

Power and planning have already emerged as weak points in Scotland’s AI infrastructure ambitions, because capital can move faster than transmission upgrades, consenting, or new generation. A funded data-centre shell is of little economic value if sufficient power cannot reach the site when the servers are ready.

DataVita has set out broader renewable-energy plans for Lanarkshire, but delivery remains subject to planning, grid, commercial, and consenting processes. Those dependencies make the construction timetable as much an energy-infrastructure problem as a data-centre one.

Growth zones face an implementation test

The UK’s AI Growth Zone policy is intended to concentrate planning support, energy coordination, and digital infrastructure in locations able to host large compute investments. Lanarkshire is among the earliest tests of whether those pieces can be aligned quickly enough to turn announcements into operating capacity.

The £300 million facility removes a major financing obstacle and gives DataVita a route to start building against contracted demand. Yet the project will still be judged on whether hardware is installed, power is available, and customers can use the capacity on schedule.

That distinction is becoming more important as governments compete to attract AI infrastructure. Access to capital can be engineered through guarantees and lender syndicates, but grid capacity, planning, and delivery capability remain harder to accelerate, and they will determine how much of the UK’s announced AI build actually reaches operation.