Data centre firms say the AI bottleneck is now power and labour
NTT Data, Brookfield, and Infosys executives cite electricity and skilled labor as key constraints for AI data centre expansion, not chip availability. European markets face longer grid connection delays. NTT Data's CEO noted labor costs and equipment delays, while Brookfield's CEO highlighted project execution as the main challenge. Infosys pointed to regulatory backlash in some regions.
How this was made
The 30-second read
Why it matters
The key market takeaway is a potential delay in AI data-centre capacity additions due to grid and workforce constraints, plus added policy risk in certain regions. This can influence valuation assumptions around buildout speed, project risk, and demand for related services.
Market read
Traders may use this as a macro-to-sector read-through for AI infrastructure buildout timelines, but it is not a fresh company-specific financial disclosure.
What to watch
The article is largely qualitative and conference-based; it does not provide company-specific contract wins/losses, so near-term trading may be driven more by broader AI infrastructure sentiment than by measurable fundamentals.
Background
Executives from NTT Data, Brookfield, and Infosys argue the AI data-centre bottleneck has moved from chips to electricity availability and skilled labour, with Europe facing especially long grid-connection timelines and some booking pauses.
Ticker impact
Brookfield Asset Management executive Connor Teskey says the bottleneck is who can pull AI data-centre projects out of the ground amid power and labour constraints.
Stock may be bid on the view that Brookfield’s execution capability matters more than chip supply.
The article provides qualitative constraints and demand-supply framing, but no Brookfield-specific deal or financial update.
Infosys executive Salil Parekh flags third risks including Western backlash and a Scotland vote to freeze new data-centre builds, affecting AI infrastructure demand.
Limited downside pressure possible if investors extrapolate slower AI infrastructure spending in affected regions.
The article discusses policy risk and hyperscale planning stages, but does not quantify Infosys revenue exposure or announce a new contract change.
Market effects
Shifts AI data-centre investment focus from chip supply to grid capacity, construction labour, and permitting, which can reprice risk for infrastructure developers and related service providers.
Europe faces longer grid-connection lead times (up to a decade or more) and policy pauses, implying uneven AI capacity growth across markets.
If power and labour constraints persist globally, AI capex may become more back-end loaded, affecting timelines for data-centre capacity and downstream IT spending.
Counterpoint
Even with power and labour constraints, long take-or-pay style commitments and strong corporate credit could keep project economics intact, limiting downside for infrastructure-linked equities.
Key entities
- companyNTT Data
CEO Abhijit Dubey says talent and electricity are now the bottlenecks, citing long equipment sourcing timelines.
- companyBrookfield Asset Management
Connor Teskey says the bottleneck is execution capacity to pull projects out of the ground, not a bubble driven by weak economics.
- companyInfosys
Salil Parekh highlights additional risks including Western backlash and a Scotland vote to freeze new builds.
- otherDenmark grid operator
Paused new connections in March with a large queued pipeline versus peak demand.
- otherScotland government
Voted to freeze all new data-centre builds, with ministers yet to decide.
