The Cloud Has a Power Problem, and It Has Reached Malaysia
On 14 July 2026, New York became the first American state to hit pause on new large AI data centers. Governor Kathy Hochul signed a one-year moratorium on any facility drawing 50 megawatts or more, citing a simple problem: these sites consume so much power that they threaten to outrun what the grid can supply, while pushing electricity bills higher for everyone else. Fourteen other states are now weighing similar moves.
This is not only an American story. In February 2026, Malaysia stopped approving new data centers that are not tied to clear AI benefits, held back by the same pressures on power and water. The country has become one of the world’s hottest data center destinations, and Johor now holds around 80 percent of national capacity. The build-out reshaping the global cloud is happening on our doorstep, and it is beginning to strain the grid it runs on.
WHAT THE RESEARCH SHOWS
-
- The New York moratorium pauses new hyperscale projects for up to a year while the state writes rules to protect the grid, ratepayers, and water supplies.
- Hyperscalers are pouring hundreds of billions into GPU-heavy AI data centers, while the older systems that run most everyday business workloads receive less investment and less maintenance.
- Forrester predicts at least two major multi-day cloud outages in 2026, caused by exactly this trade-off, as aging infrastructure falters under growing complexity.
- In response, Forrester expects at least 15 percent of enterprises to move toward private AI on private clouds this year, driven by rising costs, data lock-in, and operational risk.
- In Malaysia, the boom is concentrated in Johor, leaving the local grid and water infrastructure racing to keep up with demand.
“These incidents weren’t isolated, they’re a preview of what’s to come.”
Lee Sustar, Principal Analyst, Forrester
Why This Matters for Malaysian Businesses
Two things follow from this, and both touch the everyday running of a business. First, the cloud you depend on is being rebuilt around AI, and that transition is making it less stable, not more. When a handful of giant providers own most of the infrastructure, their upgrade cycles and their bad days become your downtime. An outage that takes a booking system, a payment gateway, or a factory’s order system offline for a day is not an inconvenience, it is lost revenue and lost trust.
Second, power and location now matter as much as price. As grids strain and governments step in, where your systems physically sit, and how many independent paths they have to stay online, becomes a real business decision. Putting everything with one distant provider is no longer the safe default it once seemed.
HOW BIGBAND HELPS
This is where a regional partner earns its place. BigBand’s role is to give you resilience you can rely on, close to where your business actually operates.
- Regional Colocation: Your own hardware in a secure Malaysian facility, close to your operations and your customers, so you are not wholly dependent on a distant hyperscaler having a good day.
- Hybrid and Private Cloud: A mix that keeps your critical workloads under your control and your costs predictable, the same shift analysts expect most resilient enterprises to make.
- Redundant Connectivity: Multiple independent network paths, so a single failure upstream does not take your business offline with it.
- Backup and Disaster Recovery: Recovery designed and tested for real outages, not just brief blips, so you keep running when a major provider cannot.
Behind all of it sits BigBand’s advisory team, helping you decide what belongs where, and keeping your infrastructure steady while the wider cloud goes through its noisiest years.
ADVISORY NOTE
Where Would Your Business Be During a Multi-Day Outage?
If your main cloud provider went dark for two or three days, could you keep serving customers? Talk to BigBand for a no-obligation resilience review, and we will map exactly where your risks sit, in plain business language.