AI Infrastructure Intelligence for Maryland

## Praevoium Intelligence Brief: Maryland AI Infrastructure Landscape **Date:** July 2026 **Executive Summary:** Maryland presents a developing, rather than fully established, AI infrastructure landscape. While proximity to federal agencies and research institutions offers long-term potential, current AI data center activity is nascent compared to neighboring Virginia. The state's power grid is robust, but interconnection queues require careful monitoring. Regulatory frameworks are generally favorable, though specific AI-focused legislation is limited. Semiconductor presence is primarily research-oriented. Maryland is currently categorized as an **Emerging Tier** for AI infrastructure investment, with potential to transition to Tier 3 within the next 3-5 years, contingent upon increased private sector investment and further development of dedicated AI compute facilities. --- ### 1. AI Data Center Activity Maryland's AI data center activity is characterized by a limited number of dedicated AI compute facilities. The state primarily hosts traditional enterprise and hyperscale data centers, with AI workloads often running on general-purpose infrastructure. * **Known Projects & Operators:** * **Digital Realty:** Operates multiple data centers in the Baltimore-Washington metropolitan area (e.g., Ashburn, VA proximity). While not exclusively AI-focused, these facilities are capable of hosting AI workloads. Specific AI-dedicated capacity within these Maryland sites is not publicly disclosed. * **Equinix:** Maintains a presence in the Baltimore area. Similar to Digital Realty, these are colocation facilities that can support AI, but are not purpose-built AI data centers. * **QTS Realty Trust:** Operates a data center in Frederick, MD. This facility offers significant capacity, but its primary focus is not explicitly AI compute. * **Hyperscale Cloud Providers (AWS, Microsoft Azure, Google Cloud):** While these providers have significant regional footprints, their primary data center campuses are predominantly located in Northern Virginia. Maryland facilities often serve as edge locations or smaller regional points of presence, rather than large-scale AI compute hubs. * **MW Capacity:** Specific MW capacity dedicated to AI within Maryland is not publicly reported. General data center capacity in Maryland is estimated to be in the hundreds of megawatts, but the proportion allocated to high-density AI compute remains a small fraction. Projections indicate a gradual increase in AI-specific rack deployments within existing facilities, driven by demand from government contractors and research institutions. --- ### 2. Power and Grid Infrastructure Maryland's power grid is managed by several utilities and is generally considered reliable. The state is part of the PJM Interconnection, a large regional transmission organization. * **Interconnection Queue:** * The PJM Interconnection queue is a critical determinant for new large-scale power consumers, including data centers. As of Q3 2023, the PJM queue contains a substantial number of projects, leading to extended interconnection timelines (projected 3-5 years for large-scale new generation or load). Data center developers in Maryland will face similar queue challenges as those in Northern Virginia, though the volume of new data center load requests in Maryland is currently lower. * Specific data center projects in the Maryland portion of the PJM queue are not individually identified in publicly available PJM reports. However, any significant new data center development would require PJM approval. * **Utility Landscape:** * **Baltimore Gas and Electric (BGE):** Serves central Maryland, including Baltimore and surrounding counties. BGE is a subsidiary of Exelon. * **Pepco:** Serves Montgomery and Prince George's counties, including areas bordering Washington D.C. Pepco is a subsidiary of Exelon. * **Potomac Edison:** Serves western Maryland. * These utilities generally have sufficient generation capacity. However, localized transmission and distribution infrastructure upgrades may be necessary for large, high-density AI data center developments. * **Renewable Energy:** Maryland has a Renewable Portfolio Standard (RPS) mandating 50% renewable electricity by 2030, with 14.5% from solar. This commitment aligns with sustainability goals for many data center operators. --- ### 3. Permitting and Regulatory Environment Maryland's permitting and regulatory environment is generally considered business-friendly, though it lacks specific legislation tailored for AI data center development. * **State Level:** * The Maryland Department of the Environment (MDE) oversees environmental permits (e.g., air quality, water discharge). * The Public Service Commission (PSC) regulates utilities and approves major infrastructure projects. * Maryland does not currently offer specific tax incentives exclusively for AI data centers, unlike some other states for general data center development. Existing incentives for general business investment may apply. * **Local Level:** * Permitting processes vary by county and municipality. Jurisdictions like Montgomery County and Prince George's County, with existing commercial and industrial zones, are generally more experienced with large-scale development. * Zoning regulations, site plan approvals, and building permits are managed at the local level. Community engagement and impact assessments are increasingly important for large projects. * **Data Privacy:** Maryland has not enacted a comprehensive state-level data privacy law akin to California's CCPA. However, federal regulations (e.g., HIPAA for healthcare data) and industry-specific compliance requirements remain applicable. --- ### 4. Semiconductor and Supply Chain Presence Maryland's semiconductor presence is primarily focused on research and development, rather than large-scale manufacturing. * **Research & Development:** * **University of Maryland, College Park:** Strong programs in electrical engineering, computer science, and materials science, with research in advanced computing architectures, AI hardware, and quantum computing. * **Johns Hopkins University:** Significant research in AI, machine learning, and specialized computing for biomedical applications. * **National Institute of Standards and Technology (NIST):** Located in Gaithersburg, MD, NIST conducts extensive research in semiconductor metrology, advanced materials, and AI standards. * **Federal Laboratories:** Proximity to federal agencies (e.g., NSA, DoD) fosters R&D in specialized computing and secure hardware. * **Manufacturing:** There are no major semiconductor fabrication plants (fabs) in Maryland. The state is not a significant hub for silicon wafer production or advanced packaging. * **Supply Chain:** Maryland benefits from its proximity to major transportation hubs (Port of Baltimore, BWI Airport) for logistics and supply chain access. However, the specialized supply chain for AI hardware (e.g., GPUs, high-bandwidth memory) is largely national or international, with distribution centers often located in larger logistics hubs outside the state. --- ### 5. Overall Infrastructure Investment Climate Maryland's infrastructure investment climate for AI is currently categorized as **Emerging Tier**. * **Emerging Tier:** This categorization reflects the following characteristics: * **Limited Dedicated AI Compute:** While traditional data centers exist, purpose-built, high-density AI compute facilities are scarce. * **Strong Research Base:** Excellent academic and federal research institutions provide a talent pipeline and innovation ecosystem. * **Proximity to Demand:** Close to federal government agencies, defense contractors, and research institutions that are significant consumers of AI. * **Robust General Infrastructure:** Reliable power grid, good transportation networks. * **Developing Policy Landscape:** General business-friendly environment, but lacking specific AI data center incentives or streamlined regulatory pathways. * **Competition from Virginia:** Significant gravitational pull from Northern Virginia's established data center ecosystem. * **Outlook:** Maryland has the foundational elements to transition to a **Tier 3** market within the next 3-5 years. This transition would require: * **Increased Private Sector Investment:** Attraction of major AI data center developers or hyperscale cloud providers to build dedicated AI compute campuses. * **Targeted Incentives:** Implementation of state or local incentives specifically designed to attract AI infrastructure. * **Workforce Development:** Continued investment in STEM education and specialized AI skills training. * **Strategic Site Selection:** Identification and preparation of shovel-ready sites with robust power and fiber connectivity. Maryland's long-term potential for AI infrastructure is significant due to its intellectual capital and government demand, but it requires deliberate strategic development to overcome its current nascent status.