By Mark Houpt, Chief Information Security Officer, DataBank
Constant Changes Require Regular Policy Reviews
Change is constant in the world of IT. The same should hold true for your security policies. Any time your IT environment changes, it’s imperative to review your security policies practices before the IT changes. This will ensure that your IT systems are aligned with corporate and general best security practices that will curtail or prevent unwanted intrusions.
Perhaps, you plan to migrate from an on-premises architecture to a colocation scenario. You might also want to set up remote users or turn to a hybrid situation that includes public and private cloud platforms. No matter which changes you contemplate, it’s time to look at your security policies. The new architecture could impact your security posture and audits for regulatory requirements. You also need to know if your enterprise can still operate within the current policies or make the necessary changes that allow the business processes to function while protecting your data and digital assets. Assure that a gap analysis is part of a security assessment in which all the policies still work effectively inside the new environment. This environment may include another data center operator (colocation or managed services provider) handling some of the workload, while other functions may remain on premises or move to a cloud platform.
Determine the Impact of Moving Data
When infrastructure changes occur, companies may not understand the impact of the new environment and what they need to change with respect to security policies and data protection.
First, determine which regulations you are required to follow, such as PCI-DSS, HIPAA, and other privacy regulations. Then run an analysis, called a Business Impact Analysis (BIA), to determine the impact of moving data into the new environment. This includes considering what measures to implement to protect data to which the new colocation or cloud provider potentially has access.
You may determine, for example, that you want to use a different encryption level. If you are subject to HIPAA, you may need to have a Business Associate agreement that addresses the activities and responsibilities of your service provider partner. Be sure to look at the regulatory and business drivers as well as the requirements within your organization — legally and within business processes — that allow or disallow the use of the new environment. For example, data sovereignty, or the location of data, is an important factor in many regulations, especially those that are assigned by a State or the Federal Government.
Collaborate with Internal and External Resources
It’s a good idea to regularly connect and interface with partners with whom you share security responsibilities. This includes your auditors, vendors (and their subcontractors), and customers to find out how the new environment will affect their audits:
● Will it impact whether you pass audits?
● Will the cost or the length of time to complete audits increase?
● Does the changing environment impact the number of documents auditors will request?
● Is there a responsibilities matrix that defines which party is responsible for each aspect of the environment?
Also, collaborate with your internal business units as well. How will their processes and data be impacted when accessing systems inside the new environment? What’s the impact on the security postures of the environments in which they operate?
Perhaps, your most important allies are your colocation and cloud providers. This involves discussing their responsibility matrix, which influences how your security policies might need to change. The matrix outlines who is responsible for implementing, managing, and maintaining specific cybersecurity controls under the Cybersecurity Maturity Model Certification (CMMC) 2.0 program and the National Institute of Standards and Technology (NIST) SP 800-171 publication.
The provider may assume some of the security responsibilities for you, or it could put new responsibilities on your plate. For example, a colocation provider or a cloud provider will take on a portion of the physical security of the data center where the data resides. However, that doesn’t remove the security requirement from you totally; you still have people accessing that data on the back end.
It’s important to identify exactly what each provider delivers for security measures and what it claims are your responsibilities. Then check what the provider’s audits say and how they perform against those audits. Are there exceptions? Are there weaknesses that might create a vulnerability within your organization?
The Team Approach to Building Strong Security Policies
DataBank consults with many customers on our colocation environments and other environments to which our data centers connect. We clearly identify our security responsibilities and make our services easy to consume, taking the worry out of IT, as much as possible, within the security requirements. We also point out where the customer is responsible and where their cloud and SaaS providers are responsible.
In some cases, we participate in group calls with our customers and their other providers to understand where each party’s responsibilities begin and end. We offer our expertise and interact in a collaborative way to help all the parties in each customer’s ecosystem ensure the implementation of strong security policies.
This is helpful when discussing the transmission of data between colocation sites and a cloud platform. A colocation provider is not responsible for securing data transmissions. However, we recommend that our customers use AES-256 encryption or better for data at rest and in transit. If the cloud provider won’t handle the encryption, we can advise the customer how to configure the encryption.
Security Policy Recommendations for Government Data
In one case, we worked with an enterprise customer that handled federal government data. One of their on-premises data centers kept crashing, so they wanted to create a hybrid environment to share the workload.
To solve this challenge, we moved part of the infrastructure inside one of our colocation data centers, and the uptime increased. We also offered our expertise on the boundary line requirements. In addition, the customer valued our interpretation of the FedRAMP and FISMA requirements for handling data and maintaining the security of that data from physical and logical perspectives.
For customers not going through a change to their infrastructure, we still recommend reviewing and updating their security policies at least annually, even if just to confirm which systems are still active. In some cases, reviews should occur more frequently. For instance, enterprises should review artificial intelligence policies for security every quarter. The technology and the overall concepts change so rapidly.
When reviewing physical security policies, the concepts haven’t changed dramatically in years. You can validate the policy on an annual basis.
Developing a Security Policy Manual
Our information security policy standards manual update involves looking at all the regulations we need to comply with to determine which one is the most reasonable and the most stringent. We set our policies across the board at the most relevant and the most business applicable model and raise the bar when a regulation or situation dictates a higher control.
That’s because we don’t want to have a negative impact on the business. The regulation we conform to might not be the most stringent, but it is the most relevant to our environment. And then we have everything else comply under that unless there’s an item that’s a unicorn.
We set our policies for the NIST 800-53 Rev. 5 moderate level. However, if there’s a situation where we need to meet high criteria, we write a policy that takes us to that level for that specific matter. This occurs with penetration testing as an example. The NIST 800-53 moderate level says we need to conduct a penetration test only once per year. But PCI-DSS compels us to run these tests twice each year. The other standards are met by the moderate baseline, and then that one is bumped up to meet the PCI-DSS requirement.
Always remember: If you just migrated to a hybrid IT environment, or anytime you make changes to your infrastructure, the time to review and change your security policies is before and after the move. One ensures that your policies are up to date and ready, the other ensures that the policies are adhered to. Otherwise, you run the risk of a breach that could lead to the loss of data or regulatory violations.
Decarbonizing data centers refers to reducing greenhouse gas emissions by improving energy efficiency, increasing renewable energy use, and optimizing infrastructure operations.
By Jenny Gerson, Senior Director of Sustainability
In today’s climate-conscious world, data center operators are playing a vital role in driving sustainable innovation. As organizations work toward net zero goals, modern data centers are stepping up, not just by delivering high-performance infrastructure, but by focusing on energy efficiency and environmental responsibility.
From greener power sources to smarter facility design, leading data center providers are taking steps to decarbonize operations. It’s a strategic shift that supports both global climate targets and the evolving sustainability priorities of investors and customers.
Understanding the Three Scopes of Emissions
Decarbonizing isn’t just about cutting carbon. It also involves knowing where emissions come from and tackling them with new approaches and more effective solutions.
For data centers, emissions fall into three categories or “scopes,” as defined by the Greenhouse Gas (GHG) Protocol. Each scope represents a different piece of their overall sustainability strategy. Most data center operators are making progress across all three.
Scope 1: Reducing On-Site Emissions at the Source
Scope 1 covers direct emissions from sources owned or controlled by the data center operator. In a typical facility, that includes backup diesel generators, on-site natural gas systems, and fugitive emissions from refrigerant leaks or HVAC systems for cooling.
Many data centers are replacing older equipment with low-emission or renewable-fuel generators, using next-generation cooling technologies, and implementing monitoring systems that detect and reduce refrigerant leaks. These actions help shrink a facility’s carbon footprint and also enhance efficiency and reliability.
Scope 2: Cleaning Up the Power Supply
Scope 2 emissions come from the electricity a data center purchases to power its infrastructure. While data centers are designed to be energy efficient, electricity is still a key part of the equation.
The good news: Data centers are leading the way in transitioning to cleaner, greener power. Many are investing in renewable energy through power purchase agreements (PPAs), renewable energy credits (RECs), and partnerships with utilities. Others are locating facilities in regions with low-carbon grid mixes or developing their own on-site solar or wind capacity. These efforts are dramatically reducing Scope 2 emissions without sacrificing performance.
Scope 3: Tackling Emissions Across the Supply Chain
Scope 3 includes all other indirect emissions – everything from the materials used to build the data center to the equipment it runs and what happens when that equipment reaches end-of-life. It also includes downstream activities, such as how tenants or cloud customers use leased infrastructure.
While Scope 3 is often the most complex to address, it’s also where data center leaders are making long-term investments that pay off. By sourcing low-carbon building materials, extending the lifespan and reuse of IT hardware, and working with vendors and partners who share sustainability goals, operators can reduce emissions. By helping customers optimize workloads and power usage, they’re creating ripple effects that benefit the entire digital ecosystem.
Together, Scopes 1, 2, and 3 offer a complete view of a data center’s carbon footprint and a roadmap for meaningful progress. With this framework in mind, let’s look at the practical steps data center operators can take today – and what DataBank is doing right now.
How DataBank is Leading with Smart, Sustainable Strategies
At DataBank, sustainability isn’t just a goal. It’s built into how we operate, grow, and serve our customers. While we’re addressing all three emission scopes, our most immediate impact comes from reducing Scope 2 emissions, starting with renewables.
By integrating sources such as solar and wind into their energy mix, data centers are shrinking their carbon footprints while maintaining the reliability and performance their customers depend on. We’re also exploring carbon capture and sequestration opportunities with utility partners, investing in innovations that reduce emissions at the source, not just in our facilities, but at the grid level. These investments align with a larger shift in how data centers interact with the energy ecosystem.
As power grids evolve, flexibility becomes just as important as efficiency. Hyperscalers are able to optimize load shifting to move power-intensive workloads like AI training to times of day when renewable energy is most available and cost-effective. Not only does this help flatten demand curves and stabilize the grid, it opens the door for more renewables to come online.
While hyperscalers are able to do this, most colocation operators can’t. Instead, DataBank engages in demand response using large-scale batteries and generators to stabilize the grid. By aligning our operations with grid dynamics, we’re not only enhancing our sustainability efforts but also contributing to a more resilient and efficient energy system.
Such a flexible approach can pay off. Recent research featured on the “The Potential for Flexible Data Centers” episode on the Catalyst podcast reveals that if data centers could be flexible – just 0.25% of the time approximately 22 hours per year – it would be possible to add up to 76 gigawatts of new data center capacity without necessitating additional grid infrastructure. This shows the significant impact that even minimal operational flexibility can have on accommodating growth sustainably.
Driving Sustainability Through Action
DataBank is making meaningful progress toward its sustainability goals by implementing a wide range of energy-efficient practices and proactive environmental initiatives.
Net zero commitment: DataBank has set a corporate goal to achieve net zero carbon emissions by 2030, focusing on reducing emissions across Scope 1 (direct emissions) and Scope 2 (indirect emissions from purchased electricity). This includes replacing fossil-fuel-based systems with more sustainable alternatives, expanding renewable energy sourcing, and exploring innovative technologies like carbon capture.
Green power usage: By the end of 2024, over 60% of DataBank’s total energy consumption (823GWh) now comes from renewable sources. This includes contractually bundled, load-following RECs and renewable power from the grid, representing a significant step toward our decarbonization goals.
Water conservation: DataBank’s use of internal closed-loop chilled-water systems in many of its data centers eliminates the need for continuous water replenishment. In 2023, the company reported zero water withdrawals at 42 of its facilities, demonstrating the effectiveness of these systems in minimizing water consumption across its operations.
EPA recognition: For the fourth consecutive year, DataBank was recognized by the U.S. Environmental Protection Agency (EPA) and ranked 54th on its National Top 100 List of the largest green power users. Additionally, the company ranked 15th on the Top 30 list of green power consumers in the Tech & Telecom sector.
Energy efficiency measures: To improve facility efficiency and reduce emissions, DataBank continues to standardize high-efficiency air-cooled technologies and deploy closed-loop chilled-water systems across its data centers. These systems both reduce energy consumption and eliminate the need for continuous water replenishment, aligning with long-term environmental sustainability goals.
Operational monitoring: DataBank tracks and reports its energy consumption, energy intensity, and Power Usage Effectiveness (PUE) to its Board of Directors quarterly and to the public annually through its Corporate Responsibility Report. This practice ensures accountability and continuous improvement in energy efficiency.
Together, these efforts demonstrate DataBank’s commitment to environmental stewardship and support its goal of building more sustainable, future-ready data centers.
Building a Greener Digital Future
As demand for digital infrastructure grows, so does the responsibility to build it sustainably. With a clear strategy and bold action across all three scopes of emissions, DataBank is not only lowering its carbon footprint; it’s helping lead the data center industry toward a cleaner, more energy-resilient future.
How to Use AI to Stop DDoS Attacks
By Tyler Treat, Director of Security Architecture
The Bad Guys Use AI: Shouldn’t You?
By debilitating IT infrastructures, DDoS attacks (Distributed Denial of Service) produce big headaches. With advances in artificial intelligence (AI), the bad guys can now strike more easily than ever.
If you don’t protect your server infrastructure properly, you will present an easy target. AI tools can quickly search for and exploit digital assets and hunt relentlessly until they find vulnerabilities.
However, you can use AI, too, to bolster your security posture and defend those same digital assets. Read on the learn how.
What Is a DDoS attack?
DDoS occurs when attackers commandeer a sizable network or pool of compromised devices to overwhelm systems, network devices, websites, or applications. For example, if a company runs a vulnerable version of a web server hosting its website, DDoS can prevent that server from responding to traffic from legitimate users.
The technique floods digital assets with malicious traffic from multiple sources: thousands or even hundreds of thousands of compromised devices known as a botnet. Unlike a denial-of-service attack that originates from a single source, DDoS is more difficult to detect, block, and trace. The attacks appear to originate from many sources.
By degrading service availability, DDoS disrupts business operations and serves as a smokescreen for other malicious activities. Sometimes, a specific service on a device may be vulnerable. If the DDoS attack hits hard enough, it can shut down the entire supporting network infrastructure.
As attacks cripple critical digital infrastructure, this leads to service outages and loss of revenue. Customers, employees, and business partners can’t access the impacted target. The volume of attack traffic also consumes internal resources and distracts security teams from addressing other threats.
Who Originates DDoS Attacks?
Traditional defenses against DDoS attacks rely on a combination of on-premises mitigation appliances, ISP-level traffic filtering, rate limiting, and cloud-based scrubbing services. These methods try to identify and block malicious traffic before it reaches critical infrastructure.
However, static rules and signature-based detection struggle to keep pace with modern, adaptive DDoS campaigns. Since DDoS attacks blend in with legitimate traffic and target application-level services, they can be difficult for legacy security tools to detect.
For enterprises that deliver high-availability applications — e-commerce platforms, SaaS providers, and colocation data centers — DDoS attacks don’t just create technical challenges. They also erode customer trust and can generate substantial financial penalties tied to customer SLAs and compliance requirements.
The person launching a DDoS attack may be politically motivated, sponsored by a nation state, or it could be a lone wolf with an axe to grind. Online services now allow people to pay in bitcoin for DDoS time and simply point an attack at a target he or she wants to hit for a specified amount of time.
DDoS activity will likely ramp up as AI tools become more mainstream. The bad guys no longer need to hire a programmer on the Dark Web who knows how to find weaknesses. AI does all the work much faster, scanning public IP addresses to find targets with vulnerabilities.
How AI Helps Defend Against DDoS Attacks
Just as cybercriminals leverage AI to launch DDoS attacks, security teams can do the same to stop DDoS attacks. AI excels at identifying subtle deviations in network behavior, enabling more precise and proactive mitigation and finding vulnerabilities missed by traditional definition-based scanners.
In addition, AI machine learning algorithms can analyze massive volumes of traffic in real time to detect anomalies that indicate DDoS events, often before traditional tools raise alerts. When integrated into security platforms, AI can then adapt to meet evolving threats. This includes reducing false positives by learning the normal traffic patterns unique to each enterprise environment.
Another plus comes from DDoS mitigation platforms that apply AI to their detection and response processes. Rather than a security engineer watching logs flow by looking for alarms, programmatic responses identify malicious traffic and then apply automated remediation measures.
This capability is key since AI can iterate attacks to keep looking for weaknesses. When security measures block an attack, AI just keeps changing identifiers.
With AI built into the mitigation platform, enterprises can monitor for new attacks and respond just as quickly to what comes next. AI acts a force-multiplier for security professionals just as it does for the bad guys.
Best Practices for Insulating Digital Assets
Run customer-facing applications on content delivery networks.
Identify high-value assets and critical services that require the most protection.
Implement layered defense measures.
Use AI tools that support behavior-based analytics, anomaly detection, and automated response.
Leverage high-quality telemetry to prioritize data hygiene and centralize logging.
Brief the InfoSec team on the decision-making process of AI tools so they trust the outputs.
Collaborating with Colocation Partners
Colocation providers offering managed security services can better protect customers targeted by DDoS attacks. The leading providers layer defenses, such as placing web servers behind load balancers with proper configurations and verifying if applications have quality web code to remove risks that could be exploited. They also offer visibility at the physical and network edge, where AI-driven detection is most effective.
When setting up a new colocation environment, collaborate with your partner to ensure integration between AI tools and the infrastructure. This includes shared telemetry feeds, synchronized incident response plans, and access to edge-layer data, such as NetFlow, syslogs, and physical access logs. This joint architecture approach enables both parties to act early on threat indicators and build a more resilient defense posture.
Another key strategy with which colocation providers can assist customers is routing Internet traffic through a content delivery network (CDN). These networks block anything going to resources at a colocation data center that’s not legitimate. A CDN can take the hit because the hosting company has a massive network and advanced scrubbing technology.
Managed Mitigation Services Block DDoS Attacks in Real Time
Here at DataBank, our colocation facilities provide managed mitigation services that eliminate DDoS attacks in real time, including stealthy sub-saturating attacks and volumetric attacks at layers 3 through 7 for both IPv4 and IPv6 traffic. Just as importantly, our service allows friendly user traffic to flow uninterrupted, even while under attack.
The DataBank DDoS Mitigation Process
As a managed services partner, we orchestrate incident response, checking if customer networks block DDoS attacks effectively, and blackholing any attacking IP addresses at upstream carriers. We analyze the style of attack as it goes through the response phases—from detection to eradication and root cause analysis.
We also offer intrusion detection and response services and apply multiple layers of defense. In addition to physical security controls, these include network perimeter controls, access controls, and host system controls. We monitor all these layers continuously to not only protect individual customer infrastructures within our colocation facilities, but also our collective customer ecosystems.
By Autumn Salama, Vice President of Cloud Operations, DataBank
Kindness in Action: A Powerful Collective Effort
At DataBank, we believe that infrastructure isn’t just about power, cooling, and connectivity—it’s about the people behind it. That’s why giving back to the communities where we operate isn’t just an initiative; it’s part of who we are. Through employee-led programs and company-supported volunteer efforts, our team regularly demonstrates that technology companies can—and should—make a meaningful difference beyond the data center walls.
Our latest effort proved just how powerful collective action can be.
On April 29th and 30th, 143 DataBankers came together—virtually and in person—as part of our Kindness in Action Campaign, led by the Women of Databank employee resource group. The mission? To give back in a big way.
Together, we assembled over 650 Kindness Kits filled with essential items, which were distributed to shelters supporting women and people experiencing homelessness in Denver, Houston, Dallas, and Salt Lake City. These kits are more than supplies—they’re small acts of care that can make a big difference.
DataBank employees volunteered at the Kindness Kits Luncheon during our Kindness in Action Campaign to support women experiencing homelessness nationwide.
A heartfelt thank-you to Yvette Hamill-Graves for sharing the powerful story of Marian House, a Baltimore-based organization that helps women rebuild their lives after incarceration by providing housing, support services, and pathways to self-sufficiency. Her presentation reminded us why this work matters so much.
(If you missed it, we encourage you to seek out Marian House’s story, especially this video—their impact is truly inspiring.)
This initiative wouldn’t have been possible without the dedication of our incredible organizers—Teresa Scott , Jodi Bell , Jessica Phlegar, Chantelle Burton , Moira McEwan , Lissa Harr, and Diane Coston —who brought our in-person events to life with passion and care.
Why This Matters
Corporate social responsibility isn’t just a buzzword at DataBank—it’s woven into our culture through employee resource groups like Women of DataBank, which empower team members to lead initiatives they’re passionate about. When employees drive these efforts, the impact is more authentic, more personal, and ultimately more meaningful for both those giving and those receiving.
The success of Kindness in Action reflects our commitment to supporting the communities that support us. As we continue growing our data center footprint across the country, we’re equally committed to growing our positive impact in each market we serve.
Thank you to everyone who participated, donated, and made this campaign possible. It’s moments like these that demonstrate what it truly means to work at a company that puts people first—not just as customers, but as neighbors, community members, and human beings deserving of dignity and support.
Thank you to everyone who attended, donated, and made this effort possible. It’s moments like these that show what it truly means to work at a company that puts people first.
By Jennifer Smith, Vice President of Sales – Mountain West
Many enterprise companies turn to DataBank for colocation services because of the level of communication we provide as well as our responsiveness to their immediate needs. One such example is a publicly traded client, which provides a cybersecurity platform that monitors data in transit between devices and cloud resources.
Provisioned from colocation data centers across the globe, the client’s platform protects thousands of customers from cyberattacks and data loss by securely connecting users and applications. Read on to hear how we helped this client open a new market.
THE BUSINESS PROBLEM: FINDING A RESPONSIVE COLOCATION PARTNER
The client wanted to expand its cloud cybersecurity services into South America. Given the speed at which the cybersecurity landscape changes, the client required a colocation partner who could provide low-latency connectivity and would respond quickly when changes are needed to the 20-cabinet infrastructure it typically deploys in colocation data centers.
At this time, the client’s platform was hosted primarily by a provider with data centers across the U.S. To serve South American customers, the client needed additional capacity in Miami. The client also needed to deploy the new build within a short timeline as customer demand in South America continued to increase rapidly.
Given the client’s dissatisfaction with the primary colocation provider in turning up new environments quickly, they decided to look for a new partner. They did not want to use the incumbent provider, even though that provider had a presence in South America.
THE DATABANK SOLUTION: A RESILIENT AND DEDICATED COLOCATION SPACE
Fortunately, the process to collaborate with a new colocation partner was already underway in the U.S. Needing better services than the incumbent could deliver, the client had started to collaborate with its technology advisory firm to identify a new colocation partner. The advisory firm recommended DataBank, a long-term partner that has successfully hosted workloads for many customers of this advisory firm.
Working with the advisory firm, the client had previously migrated workloads in Atlanta, Denver, and Dallas over to DataBank colocation facilities. These migrations launched an overall plan to eventually have DataBank host all U.S. workloads in its colocation data centers.
During a quarterly business review session, we met with the client and the advisory firm and uncovered the need to establish services in South America. Our colocation space in Miami was tight. However, we managed the space of other clients to provide this client with the resilient and dedicated caged space as well as the power and cooling its compute resources required. We also met the timeline for setting up the space by early Q4 in 2024 to deliver cloud security services to customers in South America.
THE BUSINESS RESULTS: ANOTHER REVENUE STREAM AS NEW MARKET OPENS
The level of service and the proactive communications DataBank provides are key to how well the relationship works for both the client and its advisory firm. For example, when the client needs remote hands after hours to perform physical tasks on compute resource gear, they can count on us to step right in. This helps the client streamline IT resource management and minimize internal IT team travel so they can quickly complete on-site tasks.
By expanding compute capacity with DataBank, the client has generated a new revenue stream in South America and has opened up operations on an additional continent. The client also relies on the Miami colocation data center to decrease the latency of cybersecurity services provided to customers in South Florida.
With the DataBank data center now operating at a cost-effective rate in Miami—along with data centers in Atlanta, Denver, and Dallas—the client is well on its way to establishing alternate colocation options outside of what the previous colocation provider could offer. One metro in particular that interests the client is New York City, a hub for many major customers.
As services with the incumbent expire, the client will continue to move workloads to DataBank data centers in major metros across the U.S., knowing they can achieve an ROI within a short time frame. As the client’s director of cloud services noted, “DataBank is a true partner—their team collaborates as an extension of our team. They are just as concerned about our business needs and our success as we are.”
That’s the type of relationship DataBank strives to establish with all our clients.
By: Dan Fuentes, Senior Vice President of Enterprise Sales
The computing needs of colleges and universities have evolved dramatically over the past decade. Today, the rise of artificial intelligence has only accelerated that change.
In the past, traditional academic computing once centered on simulations, data storage, and basic research models. Yet today’s higher education institutions require massive, flexible high-performance computing (HPC) resources to fuel AI-driven discovery. From training complex machine learning models to running large-scale simulations and analyzing unprecedented volumes of data, modern AI research demands HPC infrastructure, deep collaboration, and innovation on a scale few individual institutions can manage alone.
For example, earlier this year, Texas A&M announced a $45 million investment in NVIDIA’s DGX SuperPOD system. This move is expected to triple its supercomputing capacity to support AI research. This announcement follows similar moves by leading higher education institutions such as the New Jersey Institute of Technology (NJIT), the University of Maryland, and Georgia Tech. In the case of NJIT, the university expects its AI research to be one-third of its total research budget.
Several recent research findings further highlight the trend of AI and its implications for colleges and universities:
According to a 2025 EDUCAUSE survey, 57% of higher education institutions now consider AI a strategic priority. This trend is up from 49% the previous year.
However, only 13% of research institutions report being prepared to effectively harness AI technologies.
The number of AI-related bachelor’s degrees awarded in the U.S. has more than doubled since 2010, with master’s degrees increasing by 71% and PhDs by 48% over the same time frame.
AI is clearly reshaping academic priorities at every level. In a recent AI panel discussion at NJIT and sponsored by DataBank, NJIT President Teik C. Lim commented on the role it will play. “AI is creating new and extensive opportunities for innovation and knowledge creation,” he said. “AI is going to have arguably the greatest effect on the creation of knowledge, goods, and services since the invention of the internet and the smartphone.”
Higher education institutions that fail to invest in AI risk falling behind, both in research and their overall reputation.
The Business Case for AI Leadership in Higher Education
Colleges and universities are under growing pressure to embrace AI not only as a research tool but as a strategic asset for attracting top talent and funding. Institutions that can offer cutting-edge AI environments are far more likely to draw leading researchers, high-achieving students, and major grant opportunities.
In a competitive academic landscape, a university’s AI capabilities are increasingly tied to its reputation. The more advanced and “AI-ready” a campus becomes, the more it is viewed as a hub for innovation, discovery, and industry partnerships.
Beyond the technology itself, AI offers a critical business advantage for higher education. By cultivating a strong AI ecosystem, universities can secure larger research grants, generate valuable intellectual property, and foster partnerships with tech companies, government agencies, and private investors.
Weighing the Options: Build It Yourself or Partner for Speed and Scale?
As demand for AI-powered technology and workloads skyrockets, colleges and universities face a critical decision. Should they attempt to build their own data center infrastructure, or should they partner with a colocation provider to meet their needs faster and more efficiently?
At first glance, building new or expanding existing facilities may seem appealing, especially for institutions that already own disparate data centers and server rooms. Yet, the reality is far more complex and costly.
Supporting today’s high-density AI workloads demands massive upgrades in cooling systems, power capacity, water usage, and ongoing management. Higher education institutions would need to invest in not only hardware but also in advanced infrastructure like water cooling, redundant generators, and specialized environmental controls. Beyond the expense, managing such a facility would pull valuable time, funding, and focus away from their core academic mission.
Recognizing these challenges, many institutions are turning to colocation providers like DataBank. By partnering with a colocation expert, universities can get to market faster, scale their AI capabilities more easily, and keep their internal teams focused on research, education, and innovation. Instead of sinking limited funding into brick-and-mortar facilities, they can direct more resources toward the compute power, talent, and research initiatives that attract top students, researchers, and grants.
In the case of NJIT, President Lim credits the collaborative partnership they’ve had with DataBank. “DataBank helped us create a modern data center that is essential to NJIT’s AI work and aspiration,” he said. “This partnership will significantly advance the university’s strengths in AI.”
The Power of Interconnection
Beyond just providing space, power, and cooling, colocation providers deliver another critical advantage: a wide range of interconnection options and capabilities. Today’s universities operate complex hybrid environments, with research workloads spanning their own campuses, public cloud providers, hyperscalers, and even other partner institutions. To succeed, they need fast, reliable, and flexible connectivity across all these platforms. This is something traditional on-campus data centers often struggle to support.
Colocation facilities are built to meet these demands. With carrier-neutral environments and access to multiple network providers, universities gain the flexibility to design the right mix of connections for their research and operational needs. Whether linking directly to a hyperscaler or securely transferring massive datasets between campuses, colocation makes it easier to build the connected foundation that AI research and academic collaboration require.
The Future of AI in Higher Education
As AI reshapes the landscape of higher education, institutions face pivotal decisions on how to support the growing demand for AI-powered research and innovation. By partnering with colocation providers like DataBank, universities can accelerate their AI capabilities, secure critical funding, and remain competitive in an increasingly tech-driven academic environment.
