This Month's Latest Tech News in St Paul, MN - Wednesday April 30th 2025 Edition

Minnesota is rapidly solidifying its status as an AI innovation hub, propelled by robust public investment and bold infrastructure upgrades. The state legislature has allocated $2.5 million to Launch Minnesota Grants, energizing startups and advancing emerging tech through innovation grants and ecosystem-building efforts.

This wave of innovation is bringing new challenges - especially as multiple large data centers, supporting AI and cloud growth, place unprecedented demands on groundwater in suburban communities.

As detailed by MPR News' coverage of data center water usage, there's tension between economic expansion and environmental sustainability, prompting legislative proposals for stricter oversight and mandatory resource disclosure.

Community voices highlight the stakes:

“When water is taken out of one source at that rate, it will have an impact on the surrounding private wells.” - Cathy Johnson, Coalition for Responsible Data Center Development

Meanwhile, economic development experts cited in CitizenPortal.ai's analysis of Senate Bill 1832 emphasize the broader goal: to position Minnesota for long-term leadership in tech, with public-private collaboration fueling job growth and national competitiveness.

Together, these developments showcase a state reaching for the future - and wrestling with the costs of rapid progress.

Minnesota is fast emerging as a battleground for hyperscale data center investment, with tech giants like Meta and Microsoft betting big on the region's potential as an AI powerhouse.

Meta recently secured $70 million worth of land in Rosemount, signaling possible expansion beyond its already planned $800 million, 715,000-square-foot facility, as industry tax incentives and a reliable power grid attract major players to Minnesota's data center market.

Microsoft, despite pausing or slowing certain U.S. projects, remains committed to over $80 billion in global AI infrastructure investment this fiscal year, with local facilities supported by energy partnerships like Xcel Energy, highlighting both opportunity and complexity in the state for cloud and AI-driven growth.

As demand for AI training accelerates, industry analysts predict AI workloads could represent 70% of all data center demand by 2030, prompting a delicate balance between scaling up and addressing energy and sustainability challenges.

Minnesota's strategic location and supportive policies are positioning it among the nation's top secondary markets, with tabled data showing Minneapolis cited for reliable grid access and government backing:

Top Secondary Data Center Markets (U.S.)	Advantages
Minneapolis, MN	Reliable power grid, supportive government
Columbus, OH	Strategic fiber routes, affordable real estate
Kansas City, MO	Central location, strong infrastructure

“These investments, along with focused execution that delivers near-term value to our customers, will ensure we continue to lead through the cloud and AI opportunity ahead.” - Amy Hood, Microsoft CFO

For a broader industry outlook - including energy trends, megaprojects, and technology advancements - see the critical data center trends shaping 2025.

University of Minnesota researchers have achieved a major milestone in sustainable AI computing by unveiling Computational Random-Access Memory (CRAM), a hardware breakthrough that processes data entirely within memory - potentially slashing AI energy consumption by up to 2,500 times compared to existing technologies.

As the International Energy Agency projects AI energy usage could reach 1,000 terawatt-hours by 2026, rivaling Japan's national electricity consumption, CRAM's innovative use of Magnetic Tunnel Junctions (MTJs) and spintronic devices eliminates the “memory wall” seen in traditional von Neumann architectures by enabling computations directly inside memory cells, removing the energy and speed bottlenecks of data transfers.

The technology, detailed in the peer-reviewed journal npj Unconventional Computing, demonstrated results such as a 2,500x increase in energy efficiency and significant speed gains on tasks like handwritten digit recognition.

The following table summarizes key projected energy usage and CRAM's potential savings:

Year	AI Energy Usage (TWh)	CRAM Energy Savings
2022	460	Up to 2,500x less
2026	1,000 (projected)	Up to 2,500x less

“This work is the first experimental demonstration of CRAM, where the data can be processed entirely within the memory array without the need to leave the grid where a computer stores information.” - Yang Lv, lead researcher

The breakthrough positions Minnesota - and its industry partners - at the forefront of next-generation AI chip design, promising not only greener computing but also a profound reduction in resource and water consumption associated with high-performance data centers.

Explore further details in the official University of Minnesota announcement of energy-efficient AI hardware, an in-depth technical report on CRAM's performance impact, and comprehensive coverage of Minnesota's leadership in ultra-efficient AI innovation.

The rapid expansion of data centers in the St. Paul region has ignited significant debate over their water consumption, prompting calls for new regulations and greater transparency.

Minnesota lawmakers are scrutinizing proposals that would require large data centers - particularly those using over 100 million gallons annually or 250,000 gallons per day - to undergo pre-application filings, publish water usage disclosures, and finance conservation initiatives.

Community concerns have centered on potential groundwater depletion, impacts on private wells, and municipal water infrastructure strain. Legislators are weighing proposed bills such as HF3007 and HF2928, which set rigorous permitting conditions and mandate a 65% carbon-free energy threshold.

Notably, only five of 11 major projects have released initial water use estimates, with Meta's Rosemount site projecting 100,000 gallons per day at peak and CloudHQ's Chaska facility planning for up to 1.5 million gallons per day - volumes far below those of power plants, but still raising local worries.

Data Center	Location	Estimated Water Use	Notes
Meta (Facebook)	Rosemount	100,000 gal/day (peak)	Closed-loop, about 1/6 of Olympic pool
Tract	Farmington	2.4 million gal/day (peak)	Equals ~11,600 homes; max only brief
CloudHQ	Chaska	1.5 million gal/day (peak)	Plentiful city groundwater
Oppidan	Apple Valley	8 million gal/year	Minnesota Zoo uses 8x more

Legislative proposals aim to prevent unchecked withdrawals that could form groundwater “cones of depression” - lowering the water table and jeopardizing both household wells and surface-dependent trout streams.

Star Tribune analysis of Minnesota data center water consumption and Minnesota legislative bill summaries on data center regulations highlight industry resistance, citing concerns over trade secrets and investment deterrence.

Meanwhile, regional comparisons, such as those reported by Axios Twin Cities report on Midwest data center water and energy demand, reinforce the urgency for oversight as big tech targets the Midwest for its cooler climate and water abundance.

As lawmakers deliberate, the path forward will likely shape how Minnesota balances AI-driven digital growth with long-term water stewardship.

The University of St. Thomas in Minnesota is set to launch its new Master of Arts in Artificial Intelligence Leadership (MAIL) program in fall 2025, designed specifically for non-technical professionals seeking to lead organizational AI integration with a strong foundation in ethics, strategic implementation, and effective communication.

Drawing faculty and expertise from the College of Arts and Sciences, Opus College of Business, and School of Law, the MAIL program's fully online, part-time curriculum consists of 10 asynchronous courses over five terms and culminates in a capstone project portfolio.

“Graduates of the Master of Arts in Artificial Intelligence Leadership will have the theoretical, practical, and ethical background to realize the full potential of these new technologies,” said Thomas Feeney, Associate Professor of Philosophy and Program Director, underscoring the forward-looking focus on responsible AI adoption.

The flexible admissions process and employer-friendly structure make MAIL accessible for early-career professionals, managers, and executives.

For those interested in start dates and tuition, the program offers multiple entry points each year, with total tuition for 30 credits set at $27,000 and financial aid options available.

The following table outlines the program's structure and costs:

Feature	Detail
Start Dates (2025-2026)	Sept 3, Oct 23, Feb 2, Mar 23, May 27, July 13
Program Delivery	10 online courses over 2 years (asynchronous)
Total Tuition	$27,000 (30 credits at $900/credit)

For in-depth program details, sample course plans, and to register for an upcoming virtual info session, visit the University of St. Thomas Master of Arts in Artificial Intelligence Leadership official announcement, review the comprehensive MAIL program overview on the College of Arts and Sciences website, and find admissions and tuition information on the University of St. Thomas MAIL admissions and tuition page.

Federal cuts to the Manufacturing Extension Partnership (MEP) may soon devastate Enterprise Minnesota's ability to support rural manufacturers, as $3.2 million in annual federal funding - representing 45% of its budget - faces elimination and would shrink service coverage from 82 counties to just 12.

This funding underpins critical growth and certification initiatives for over 500 small rural manufacturers, who together generated $696 million in new revenue and added 6,600 jobs over five years.

Industry and political leaders, including Senator Amy Klobuchar and Congresswoman Chellie Pingree, are urgently lobbying for program restoration, pointing out that the MEP network nationwide returns over $17 in economic benefits for every federal dollar invested and has created or retained more than 1.6 million jobs since 1988.

A sudden administrative move to withhold funds from 10 states prompted widespread industry outcry and a temporary funding reinstatement through the end of 2025, but future support remains at risk as federal priorities shift toward emerging technologies like AI and quantum computing.

As Bob Kill, CEO of Enterprise Minnesota, warns,

“We're going to fill [congressional members'] inbox with emails and voicemails from manufacturers calling. We've moved to that point. It's urgent. It's terrible.”

The following table summarizes the recent impact of MEP funding:

MEP Program Metric (FY2024/5 Years)	Value
Minnesota – Rural manufacturers reached	510+
Minnesota – New jobs created	6,600
Minnesota – Added revenue	$696 million
Nationwide – Jobs created/retained (FY2024)	108,000
Nationwide – Annual federal ROI	17:1

Read more about the local ramifications in the Star Tribune's in-depth analysis of Enterprise Minnesota's federal funding threat, explore the broader policy debate and industry reaction in IndustryWeek's coverage of shifting MEP appropriations, and see the wider national implications in UIC's opinion piece on the risks of dismantling the MEP program.

X Corp. (formerly Twitter) has filed a federal lawsuit challenging Minnesota's 2023 law banning the dissemination of election-related deepfakes, igniting a debate over free speech, platform liability, and the rise of AI-manipulated content.

The law makes it a crime to knowingly share highly realistic, AI-generated videos, images, or audio with the intent to harm a candidate or influence an election within 90 days before voting or party conventions, carrying penalties of up to 90 days in jail for a first offense and up to 5 years for repeat violations.

X Corp. argues this legislation is “vague and unintelligible,” forcing social networks to err on the side of broad censorship - including parody or satire - to avoid potential prosecution.

As Professor Alan Rozenshtein of the University of Minnesota notes,

“I think this lawsuit is actually quite strong and that this law...is very likely to be struck down on both First Amendment grounds and also federal statutory grounds.”

Minnesota officials, however, assert the law is precisely targeted at protecting elections from misinformation, with State Senator Erin Maye Quade saying the law prevents “spreading deepfakes meant to harm candidates and influence elections.” The dispute highlights broader national trends, as at least 24 states have enacted or are considering similar regulations.

For a full review of the legal arguments and content policy implications, see AP News analysis of X Corp.'s lawsuit against Minnesota's deepfake law, Courthouse News on legal context and enforcement penalties of election misinformation laws, and St. Cloud Times coverage of Minnesota's political deepfakes legislation.

Key Elements	Minnesota Deepfake Law
Applies Within	90 days before election/convention
Targeted Content	Realistic AI-generated images, audio, video
Intent Required	To injure a candidate or influence election
Penalties	Up to 5 years in prison for repeat offenses

KPMG's Minneapolis office is part of a nationwide acceleration of AI-powered internal audit and risk management, as the firm continues integrating advanced AI agents into its global audit platform, Clara.

These agents are designed to automate routine audit tasks - including expense reviews and the search for unrecorded liabilities - freeing over 95,000 auditors worldwide to focus on high-risk and complex issues, thus improving audit quality and operational efficiency.

The Clara platform's “Financial Report Analyzer” AI engine assists auditors in completing disclosure checklists, while ongoing upgrades promise even deeper insights with human oversight.

As Scott Flynn, KPMG US Vice Chair of Audit, explains,

“We're continuing to build out our AI capabilities with increasingly sophisticated agents in KPMG Clara to enable our auditors to more effectively respond to risks and deliver deeper audit insights. As KPMG firms accelerate the adoption of this innovative technology and deploy new capabilities, we are maintaining a high level of professional skepticism and upskilling KPMG professionals to drive trust in the capital markets.”

Recent survey findings highlight tangible business results: 93% of leaders report AI investments have strengthened competitiveness and long-term strategic performance, and 97% see measurable profitability.

The table below summarizes key statistics on AI adoption across U.S. finance and audit functions:

Metric	Value
AI use in finance functions (moderate/extensive)	62%
GenAI embedded in workflows	35%
Organizations piloting AI-agent programs	65%

The transformative impact and practical applications of these tools are discussed in depth in How AI and Data Analytics Can Transform Internal Audits, while a technical dive into the Clara upgrade's features - including autonomous risk analysis and enhanced productivity - is available at KPMG Advances AI Integration in KPMG Clara Smart Audit Platform.

For survey insights on emerging trends in AI-driven risk management, see the KPMG AI Quarterly Pulse Survey.

As Minnesota emerges as a hotspot for AI-powered data centers, concerns over their substantial energy and water footprints are fueling stringent legislative proposals and community debate.

New bills, such as HF2928 and HF3007, seek to regulate environmental impacts by requiring pre-application scrutiny for facilities surpassing 100 million gallons of annual water use or 250,000 gallons daily, mandating environmental impact statements, and imposing carbon-free energy sourcing targets of 65% for large operations.

Public disclosure of energy and water consumption is also at the forefront, though industry groups warn that such transparency poses security risks and may deter investment.

Data centers - operated by tech giants like Amazon, Meta, and Microsoft - will primarily draw from groundwater, raising local worries about aquifer depletion, private well reliability, and the needs of trout streams fed by groundwater springs.

Notably, while water usage for cooling servers varies by technology and is dwarfed by the sector's electricity demands, the table below shows estimates for key Minnesota projects:

Data Center	Location	Estimated Peak Water Use	Notes
Meta (Facebook)	Rosemount	100,000 gallons/day	Closed-loop; approx. 1/6 Olympic pool/day
Tract	Farmington	2.4 million gallons/day	~11,600 homes' daily use; brief peaks
CloudHQ	Chaska	1.5 million gallons/day	Served by city groundwater; capacity available
Oppidan	Apple Valley	8 million gallons/year	MN Zoo uses 8x more per year

Policy efforts are increasingly oriented toward ensuring data centers contribute to conservation initiatives for low-income households, while also protecting Minnesota's water resources and keeping residential utility costs in check.

As one resident reflected at a public forum,

“What's that going to do to the people that have wells?”

For a detailed breakdown of the evolving regulations and local perspectives, see the Session Daily overview of proposed water usage bills, the Star Tribune's data center water consumption analysis, and the GovTech coverage of Minnesota's data center debate.

Researchers at the University of Minnesota Twin Cities have achieved a major leap in artificial intelligence efficiency by experimentally demonstrating a magnetic tunnel junction (MTJ)-based hardware device that reduces AI's energy consumption by up to 1,000 times compared to conventional approaches.

This cutting-edge device uses Computational Random-Access Memory (CRAM), processing data directly within memory to eliminate the energy-draining transfers between logic and memory prevalent in traditional Von Neumann architectures - a shift made possible by leveraging spintronics, which utilizes electron spin rather than charge.

One experimental highlight showed savings up to 2,500 times over legacy systems, paving the way for sustainable, high-performance AI. As spintronics gains ground, its remarkable qualities - such as non-volatility, higher integration density, and rapid computation - position Minnesota at the forefront of next-generation chip technology and neuromorphic computing.

The International Energy Agency forecasts AI's global energy demand to double to 1,000 terawatt-hours by 2026, underscoring the breakthrough's significance for both environmental and industry impact.

Metric	Value
Energy Reduction	1,000–2,500× (vs. traditional)
Current AI Energy Use	460 TWh (2022)
Projected AI Energy Use	1,000 TWh (2026)

"CRAM is very flexible and more energy-efficient than traditional AI system building blocks because computation happens where the memory is."

To learn more about MTJ devices and UMN's patented advances, read the University of Minnesota's official announcement on their state-of-the-art device, discover how CRAM technology addresses AI's energy challenges, and explore the foundational role of spintronics in next-gen electronics.

Minnesota is quickly emerging as a standout player in the Midwest's surging data center sector, as tech giants and hyperscalers seek secondary markets offering cooler climates, abundant water sources, and access to affordable land and energy.

While primary data center hubs like Chicago and Columbus remain central, Minneapolis and the broader Twin Cities area are being targeted for expansion due to rising demand from the artificial intelligence (AI) boom, which powered a record $57 billion in global data center investments in 2024 and drove North American supply growth by over 40% year-over-year.

This transformation comes with challenges: vacancy rates have dropped to as low as 3–4.9% in major U.S. markets, most new capacity is pre-leased, and the environmental impact - including energy and water strain - prompted new oversight measures and legislative debates in Minnesota.

As developers rapidly secure large-acre, high-power-capacity rural sites, states across the Midwest offer data-center-specific tax credits and utilities scramble to build out infrastructure or partner with providers of renewable and nuclear power.

The table below highlights key statistics from recent market research:

Region/City	Operational Capacity (GW)	Vacancy Rate	2024 Investment/Absorption
Virginia	13.2	<1%	$57B (global), 1.5GW absorbed
Chicago	2.4	4.9%	80%+ pre-leased
Minneapolis	Emerging	N/A	Rapid pipeline growth

Regulatory scrutiny is intensifying alongside expansion, as water and energy consumption data often remain undisclosed, and lawmakers fear rising utility bills for residents.

As one Axios report outlines Midwest data center growth and impact on energy usage, experts urge transparency and next-generation technologies to keep Minnesota's ascent both economically and environmentally sustainable.

For a broader look at how AI's infrastructure boom is shaping new development frontiers, see this in-depth analysis of global data center investment driven by AI.

Lastly, as detailed in Cushman & Wakefield's 2025 Americas Data Center Update report, the rush to secure power, land, and cutting-edge cooling in emerging hubs like Minneapolis marks a defining opportunity for Minnesota to become the Midwest's next AI and data infrastructure powerhouse.

As Minnesota leans into the rapid expansion of artificial intelligence, the state is balancing innovation with a strong commitment to responsible AI growth. State agencies are leading by example, deploying secure generative AI tools like Copilot Chat within Microsoft 365 Government, with Minnesota IT Services emphasizing,

"Our commitment to innovation goes hand in hand with our responsibility to safeguard Minnesotans' data, ensuring that AI is used securely and effectively to enhance public services."

This comes as the policy landscape shifts at the federal level, following deregulatory executive orders encouraging faster AI adoption yet raising concerns over ethics and public trust, especially in sectors like healthcare and youth technology use with new state-level secure AI integration.

Efforts in St. Paul and beyond include crafting legislative safeguards to combat AI misuse, notably in protecting against deepfakes and deceptive technologies targeting children, echoing findings in the state's latest report on youth well-being and emerging tech from the Minnesota Attorney General.

Meanwhile, national momentum grows - over 550 AI-related bills were introduced across 45 states this year to address risks such as privacy, bias, and consumer protection.

Minnesota's proactive approach, reflected in collaborative workshops on cybersecurity and responsible AI, positions it as a leader in integrating ethical AI while building public trust.

For an overview of this legislative wave and how Minnesota fits into broader state actions, consult the National Conference of State Legislatures' 2025 summary of AI legislation:

Focus Area	Recent Action in Minnesota
Secure AI Deployment	Rollout of Copilot Chat for state agencies
Youth AI Impacts	AG's report urges guardrails for youth safety
AI Legislation	550+ bills introduced in 45 states; proactive local policies