Rhode Island is expanding AI training while confronting unresolved questions about data-center power costs, tax incentives and permanent jobs. Smithfield’s proposed corporate complex offers the clearest local test.
Rhode Island has moved to train workers for artificial intelligence, build a state policy apparatus around it and court related investment while lawmakers and one town confront the power, water and tax questions raised by data centers, a test of whether the technology’s economic promise can be converted into measurable public benefits.
The state’s AI push began formally in 2024, when Gov. Dan McKee created an AI Task Force and directed state agencies to develop training and education opportunities for public employees, students and private-sector workers. The order also called for AI and data centers of excellence within state government.
Rhode Island’s AI Action Plan, dated Jan. 26, recommends short-form credentials, registered apprenticeships, training incentives, public workshops and a physical and online AI-RI hub. It identifies education, defense and maritime technology, life sciences and cross-sector collaboration as areas where the state could compete.
The plan is largely a roadmap, setting out recommendations rather than a completed budget or statewide targets for enrollment, job placement and investment returns. That distinction is becoming more consequential as AI policy moves from planning documents into university programs, utility regulation, tax policy and local land-use disputes.
Matt Stoller argued this week in his BIG newsletter that warnings about catastrophic AI risks can benefit large technology companies by encouraging rules that slow competition and protect incumbents. He contends that the focus on existential danger can divert attention from conventional responses to corporate misconduct, including product liability, criminal law and antitrust enforcement.
That argument is commentary rather than a finding about Rhode Island’s programs. The state’s records instead show officials preparing for potential workplace disruption while promoting AI as an economic and educational opportunity.
A draft 2026 workforce plan from the Governor’s Workforce Board describes an analysis conducted with Brown University researchers of occupations with high employment and high exposure to AI. The plan treats that analysis as an early warning system and calls for AI literacy, career pathways and support for workers who may be vulnerable to disruption.
Exposure is not the same as displacement. The U.S. Bureau of Labor Statistics defines occupational exposure as the extent to which AI could assist with or complete tasks, rather than a measurement of job losses, automation probability, productivity gains or wages.
The clearest programs are emerging in higher education. The University of Rhode Island announced an undergraduate certificate in AI and machine-learning engineering and an AI and business analytics certificate for the 2026-27 academic year.
URI has also described workshops intended to prepare participants for AI-related employment and an AI lab that mentors approximately 35 students each semester. Those programs establish a training pipeline, while statewide placement and wage results will depend on how employers use the credentials.
National adoption data suggest that use is growing but remains uneven. The Census Bureau found that 17% to 20% of U.S. businesses reported using AI during surveys conducted from December 2025 through May, compared with 37% of companies employing at least 250 people.
Unlike software training, data centers require large physical investments in electricity, land and cooling. That has pushed Rhode Island’s AI debate toward the more familiar questions of who receives a tax break, who pays for infrastructure and what a project provides in permanent employment.
A January General Assembly release said Rhode Island had seven data centers operated by four providers. The count did not classify the facilities by whether they supported AI or provide their combined electricity and water use.
Lawmakers proposed two approaches this year. One would have offered tax incentives for qualified data centers, while two other bills sought to prevent facilities with large power demands from shifting attributable infrastructure costs to other electric customers.
House bill H7331, introduced on Jan. 28 and referred to the House Corporations Committee, would apply to data centers with projected or actual demand of at least 50 megawatts. It proposed requiring those facilities to bear costs reasonably attributable to generation, transmission or distribution infrastructure needed to serve them.
The bill also called for annual disclosures covering water consumption, cooling technology and water-recycling practices, along with potential efficiency plans and assurances that a site could be restored. Senate bill S2776, introduced on March 4, contained similar electric-cost provisions for facilities meeting the 50-megawatt threshold.
A separate proposal, S2346, would have created data-center tax incentives. The Rhode Island Business Coalition supported the measure as a way to compete with neighboring states for investment, while Green Energy Consumers Alliance warned that exemptions could erode the tax base and place pressure on other taxpayers and ratepayers.
Good Jobs First, a subsidy watchdog group, told lawmakers that data-center exemptions can be open-ended, difficult to track and weakly connected to permanent employment. Its testimony relied on projects outside Rhode Island rather than a fiscal estimate for a development in the state.
Rhode Island Sens. Jack Reed and Sheldon Whitehouse raised the utility issue at the regional level in January, asking ISO-New England how it would protect residential customers and grid reliability as data-center demand grows. They urged the regional grid operator to ensure that technology companies bear the costs created by new loads rather than shifting them to households.
Smithfield has become the most immediate local test. Town records show that Hanton City Investments LLC presented a business-park concept in April involving possible data-storage centers, followed by a formal Smithfield Corporate Complex application received on May 5.
The town’s planning repository identifies the proposal as a major land-development master-plan application. Hanton City Investments and ROK LLC later appealed a second certificate of incompleteness issued on June 11.
Publicly documented terms for the proposed complex do not identify a confirmed data-center tenant, electricity demand, cooling system, water consumption, construction cost, public subsidy or permanent-job commitment. Those figures would determine whether the proposal resembles a conventional business park, a large utility customer or something in between.
The Smithfield proposal followed a separate March 19 Planning Board discussion of a zoning amendment that would define data centers and list them as prohibited in every zoning district. At an April 15 Town Council work session, the developer had not yet obtained permits or appeared before the planning or zoning boards.
The state’s economic argument rests for now on preparation: create credentials, teach AI literacy, develop an institutional hub and position Rhode Island industries to use the technology. Its protective approach rests on thresholds and disclosures intended to assign infrastructure costs and reveal resource use before large facilities are established.
Nationally, Census Bureau data show data-center employment increased from 306,000 in 2016 to 501,000 in 2023. In Rhode Island, URI’s newest AI and business-analytics certificates began with the 2026-27 academic year, alongside the university lab mentoring approximately 35 students per semester.