The artificial intelligence fever has skyrocketed demand for hyperscale data centers, but New York State has decided to hit the brakes. Governor Kathy Hochul signed an executive order imposing a moratorium of up to one year on granting environmental permits for new facilities, while a regulatory framework is developed to protect taxpayers, the power grid, the environment, and local communities. This measure, described as the first of its kind in the country, could mark a turning point in how these critical infrastructures are regulated.

A Necessary Pause Amid 'Unprecedented' Demand

New York, like other states, faces 'unprecedented' demand for data center development that would require 'enormous amounts' of energy and water, according to Hochul. Local communities' reactions are accelerating at the same pace, and the moratorium aims to buy time to assess impacts and establish clear rules. Matt Kimball, vice president and principal analyst of data center technologies at Moor Insights & Strategy, describes this as 'a symptom of a larger national problem: computing demand far exceeds the capacity of the power grid.'

What Does the Moratorium Entail?

During the one-year pause, New York will develop a 'Generic Environmental Impact Statement' (GEIS) to assess the effects of data center construction and operation, including water consumption, energy use, and air quality. Additionally, within 60 days, a Community Investment Framework (CIF) will be published to guide local governments in negotiations with operators. This framework proposes a contribution of $1 million per megawatt (MW) of expected demand, meaning a 50 MW center would need to reinvest $50 million into the host community.

The CIF will include 'good neighbor commitments' regarding landscaping, noise and light mitigation, labor commitments with unions, and a community investment fund to support local economic vitality. Operators could fund public infrastructure, housing, job training, or broadband expansion. 'Having a published playbook on how to make this work statewide, rather than having to negotiate county by county, should be an advantage for everyone,' says Kimball.

The End of Data Center Subsidies

One of the most notable aspects of the measure is the potential repeal of sales tax exemptions for existing large data centers. Kimball emphasizes that 'states across the country have been subsidizing expansions for years to attract data center business. This could be the beginning of the end for those subsidies in many states.' The moratorium also contemplates creating a fund for data centers to invest in the state's aging power grid and support the acquisition of new clean energy.

Impact on IT Companies and Professionals

For IT leaders, this news signals that power supply and permitting are now 'first-order variables' in infrastructure strategies, alongside cost and latency. Companies basing their cloud strategy on hyperlocal capacity must reassess their assumptions. CIOs and IT leaders should collaborate with providers to gain clarity on regional capacity. The moratorium could trigger a 'border jump' to neighboring states like Pennsylvania, Connecticut, or New Jersey, but it won't be widespread. Instead, gradual pressure is expected: lower colocation availability and firmer pricing in the New York metro area, while cloud providers may direct new AI capacity to regions like Georgia, Ohio, Texas, and Utah, where power supply and permitting are more predictable.

Jeremy Roberts, senior research director and content manager at Info-Tech Research Group, considers the moratorium a 'turning point' and a way to appease an outraged public. 'People don't like that data centers don't generate many jobs, take up a lot of space, consume a lot of energy and resources, and can be noisy and smelly.' However, the impact will be macro, as the moratorium focuses on hyperscalers. 'If you were planning to build a data center in New York and your name isn't Satya Nadella, this probably won't affect you,' Roberts jokes.

Consequences for Hardware and Innovation

Roberts points out that companies are diverting IT budgets from software and mainframes to accumulate AI hardware, such as servers and memory chips, anticipating supply issues and price hikes. This is reflected in the historic drop in IBM's stock. 'If companies plan to invest in anything that uses storage or CPU, they will pay more in the future,' he warns. However, constraints often inspire innovation: hyperscalers could chain together smaller data centers or adapt in other ways. 'People are infinitely creative,' Roberts concludes.

For companies looking to optimize their operations, having tools that allow precise resource control is crucial. For example, a time tracking and clock-in system can help record work time per project, improving efficiency. Likewise, continuous training is key; an e-learning module allows you to launch and manage online courses to keep your team updated. Infrastructure security is also a priority: configuring secure VPNs and firewalls is essential to protect your network. And for businesses with specific needs, a CRM for workshops and SAT facilitates work order management, parts stock, and customer portal.

Original source: ComputerWorld. Analysis and adaptation by ForgeNEX.