NASDAQ: CRWV

We deliver our cloud services through Infrastructure Services, Managed Software Services, and Application Software Services, including proprietary storage solutions, CoreWeave Mission Control, as well as developer tools, all critical to unlocking continued innovation of AI products. Our storage capabilities are designed to ingest, move and serve large volumes of data directly into AI compute environments with the throughput and access patterns required by modern training and inference workloads. Our software solutions, including Weights & Biases®, provide AI researchers and developers with the tools required to train and fine-tune models and to build, deploy, and scale the applications through which AI is delivered to end users. These services operate as a single stack, enabling customers to provision resources, manage workloads, and monitor performance across their AI environments through a unified operational layer.

Security is a foundational component of our platform architecture and operations. We design our infrastructure, software, and operational processes to support secure isolation of workloads, protect data throughout its lifecycle, and enable customers to deploy AI workloads with consistent security controls across environments.

We operate our platform across a distributed network of purpose-built data centers interconnected through a proprietary global network backbone and largely located near major metropolitan areas. Our data centers are designed to support high-density AI infrastructure and incorporate advanced networking equipment, enhanced access to power, and, where appropriate, liquid cooling technologies. Our low-latency data center footprint and integrated software platform support fungible use of infrastructure across training and inference workloads, enabling customers to transition capacity seamlessly as their workload requirements evolve.

We have scaled rapidly over the last several years. As of December 31, 2023, we operated 10 data centers with approximately 70 MW of active power. As of December 31, 2024, we operated 32 data centers with approximately 360 MW of active power. As of December 31, 2025, we operated 43 data centers with over 850 MW of active power. As of December 31, 2025, our total contracted power capacity was approximately 3.1 GW, which we expect to deploy over future periods.

We maintain relationships with semiconductor manufacturers, original equipment manufacturers ("OEMs"), original design manufactures ("ODMs"), and software providers to support the sourcing and deployment of infrastructure components used in our platform. We primarily finance our infrastructure development through asset-level debt supported by take-or-pay customer contracts, supplemented by corporate-level equity and debt financing.

5

Table of Contents

We serve a range of customers, including leading enterprises, AI labs, and technology companies deploying AI at scale across a variety of use cases. Customers generally access our platform through multi-year committed contracts, under which they purchase a specified amount of capacity on a take-or-pay basis over the contract term. A number of our customers have expanded their initial commitments over time by contracting for additional capacity as their AI workloads have grown. We also offer on-demand access through a pay-as-you-go model. As of December 31, 2025, we had $60.7 billion of remaining performance obligations ("RPO"), compared to $15.1 billion of RPO as of December 31, 2024. As of December 31, 2025, our committed contracts had a weighted-average contract duration of approximately five years.

Our business has grown rapidly since inception. Our revenue was $5.1 billion, $1.9 billion, and $229 million for the years ended December 31, 2025, 2024, and 2023, respectively. During these periods, we continued to invest significantly in expanding our infrastructure and operations. As a result, we incurred net losses of $1.2 billion, $863 million, and $594 million for the years ended December 31, 2025, 2024, and 2023, respectively.

CoreWeave Cloud Platform and Product Offerings

A Comprehensive AI-Native Stack Powers Our Purpose-Built AI Cloud

CoreWeave is the Essential Cloud for AI, purpose-built to accelerate breakthroughs by AI pioneers, from leading research labs to enterprises fueling business growth. Our customers rely on CoreWeave Cloud for their demanding use cases—including AI model training, agentic AI and inference, agent development, and specialized workloads—because it provides the pace, performance, and transformative partnership they need.

CoreWeave has built a comprehensive, tightly integrated platform of advanced infrastructure and proprietary software that enables customers to build and run AI systems continuously, reliably, and at global scale. Our focus on AI workloads enables us to deliver best-in-class services, solutions, and infrastructure that exceed the performance and capabilities of general-purpose clouds.

We integrate security throughout our AI-native cloud, so it underpins the systems, hardware, and architecture that keep AI infrastructure reliable, transparent, and resilient. We implement top-tier industry best practices to prioritize security and privacy at every level—identity and access management, Kubernetes and container security, storage encryption, and network access controls and encryption.

Recognizing our leadership in performance, reliability, and large-scale AI cloud solutions, CoreWeave Cloud achieved its second SemiAnalysis' Platinum ClusterMAX™ rating in 2025. CoreWeave remains the industry's sole Platinum provider. We continue to iterate on CoreWeave Cloud via innovation and acquisition—enhancing its capabilities and increasing the value it brings to our customers.

CoreWeave Cloud—built from the ground up for AI innovation

Every layer of our stack is purpose-built for AI innovation and optimized for the performance, agility, and intelligence demanded by frontier-scale AI. Security is engineered into every layer of CoreWeave Cloud, from hardware through orchestration. It keeps protection and performance aligned so every new advance in AI can happen with speed, integrity, and reliability.

Networking Backbone. Our network backbone supports the bandwidth and connectivity that our customers need to move large-scale distributed AI workloads. We designed and built a carrier-grade network backbone engineered specifically for the ultra-low latency, high-throughput requirements of our customers' training, data replication, inference, and agentic AI initiatives. Our state-of-the-art backbone connects data centers across North America and Europe with high speed terrestrial and subsea fiber optic cables designed for massive scalability, reliability, and security. CoreWeave's networking backbone is designed with direct peering to major carriers and provides hybrid and multicloud connectivity, supported by our "Zero Egress Migration" program that can reduce or eliminate data transfer or egress fees. Customers can connect to our backbone via high-speed Direct Connect tailored to a wide range of requirements—with physical connectivity options of 10Gbps, 100 Gbps, and 400 Gbps of bandwidth. We continue to enhance our networking backbone to serve the evolving business needs of our customers, including large AI labs leveraging our backbone to connect multi-cloud inference infrastructures.

Foundational Infrastructure. Our purpose-built foundational infrastructure maximizes performance with first-to-market Graphics Processing Unit ("GPU") clusters, ultra-high density, and high-speed interconnects, supporting complex AI workloads while improving efficiency and lowering total cost of ownership. Our track record and ability to

6

Table of Contents

deploy the industry's most advanced hardware and architectures first—including NVIDIA GB200 and GB300 NVL72 systems for mission-critical AI—gives our customers a measurable edge in performance, efficiency, and scale. CoreWeave clusters also use cutting edge CPUs, including AMD and Intel chips, to power compute-intensive projects and to help our customers get more out of GPU compute.

Our data center fabric is designed to harness the full potential of each GPU and to maximize performance from the newest architectures. It integrates state-of-the-art networking technology, including NVIDIA Quantum-X800(XDR) enabling 800Gbps of connectivity, NVIDIA Quantum-2 InfiniBand interconnect and NVIDIA Spectrum-X with RoCE or RDMA over Converged Ethernet for ultra-low-latency connectivity between racks of GPUs. To ensure sustainability, many of CoreWeave's data centers draw on renewable energy, deploy advanced closed-loop liquid cooling systems, and pilot innovative projects—such as heat recapture to supply excess heat to nearby homes and industrial facilities.

This powerful, proven infrastructure enables exceptional results for our customers. In recent benchmarking through NVIDIA's Exemplar Cloud program, we demonstrated improvements on system model flop utilization (“MFU”) over reference targets for both training and inference workloads. For training, our infrastructure enables superior utilization compared to reference targets. For inference, our platform met or exceeded benchmark tests across inference tasks, such as reasoning and generation, demonstrating CoreWeave's consistent performance and faster scaling for every stage of AI deployment.

We continue to extend and enhance our foundational infrastructure with the latest GPU and CPU innovations, and expect to be among the first cloud providers to deploy the NVIDIA Rubin platform, expanding our support for large-scale inference, reasoning, and agentic AI.

Data and Storage. CoreWeave's purpose-built storage combines exascale, AI-optimized object and file storage with GPU-local caching and Local Tier Acceleration (LOTATM) to deliver high-throughput data access, cross-cloud reach, and predictable economics for training and inference. Our customers rely on our fast, resilient, and easy-to-use storage solutions to train, fine-tune, and deploy models more rapidly at scale—while reducing replication costs, egress fees, or performance tradeoffs. Our Local Object Transport Accelerator ("LOTA") brings data closer to GPUs by caching directly on GPU nodes, which enables rapid throughput and high uptime and durability. CoreWeave object and file storage services are optimized for AI, enabling streamlined training, low latency, efficiently utilized clusters, and faster time to market. Our microservices-based architecture allows us to support the storage solution that our customers choose.

Infrastructure Control. Our integrated AI-native orchestration and bare-metal control deliver the reliability, flexibility, and efficiency required to run complex AI workloads at scale, and matched with the right AI services. CoreWeave Kubernetes Service ("CKS"), our managed Kubernetes service, minimizes the burden of managing large GPU clusters. CKS combines fast performance, security, and flexibility in a fully managed Kubernetes solution—one that reduces management time for complex Kubernetes clusters. CKS clusters leverage bare-metal performance to gain higher GPU cluster performance, accelerating model download speeds and spin up times for inference. CKS runs on CoreWeave Cloud’s bare-metal GPU infrastructure, eliminating hypervisor overhead to maximize performance, including high-throughput model downloads and faster container spin-ups for inference. CKS clusters leverage Bare Metal nodes without a hypervisor to maximize node performance. Our data center architecture uses Data Processing Units to enforce tenant isolation and strengthen the security boundary between workloads. It also offloads processes such as networking and security to enable peak GPU performance while providing complete isolation and acceleration with private-cluster Virtual Private Clouds.

CoreWeave Mission Control. CoreWeave Mission Control integrates security, observability, and services—including node, rack, and fleet lifecycle management—to enable intelligent, unified orchestration from foundational infrastructure to agent development. It connects all of the critical layers of CoreWeave, giving customers real-time visibility into GPU, network, and storage behavior. It integrates secure controls for AI workloads, continuous operational insight, and proactive remediation paths.

With CoreWeave Mission Control's proprietary monitoring and management solutions, users can see how systems are performing, spotting early warning signs that enable them to replace or repair problematic nodes before they create delays—keeping systems stable and projects on schedule. Combining identity and access controls, compliance logging, and audit history, CoreWeave Mission Control provides a complete, clear, and defensible record of activity across an environment. Its continuous operational insight delivers deep knowledge of complex environments. Audit and telemetry signals stream seamlessly into security information and event management systems on any cloud—along with health checks on GPUs, nodes, and racks, so our customers always know the state of the system in real time.

7

Table of Contents

CoreWeave Mission Control's proactive remediation paths transform insights into action, enabling users to anticipate issues (including potential node failures) before they happen. It initiates the right responses, from automated recovery to routing the incident directly to CoreWeave experts, so customers no longer have to waste valuable time chasing ambiguous alerts or guessing at root causes. We continue to add new CoreWeave Mission Control features—including Telemetry Relay, GPU Straggler Detection, and the CoreWeave Mission Control Agent—that bring improved visibility and control to large-scale AI workloads.

Runtime Acceleration. CoreWeave delivers platform services that accelerate training and inference by optimizing workload latency, improving throughput, and increasing utilization at runtime—removing friction across scheduling, environment readiness, and execution so jobs reach steady-state performance faster and run more efficiently at scale. SUNK (Slurm on Kubernetes) helps customers deploy and operate large-scale Slurm-based AI research clusters on CoreWeave Cloud, leveraging our purpose-built infrastructure and integrated platform capabilities to support high-performance AI workloads. With a Slurm cluster purpose-built for AI, researchers can save valuable time and speed critical breakthroughs. They can run training, inference, and reinforcement learning ("RL") on a single high-performance cluster built for research at massive scale—all with automated user access management and deep observability. Our customers get metal to model visibility, resolving issues quickly, raising quality, and enabling rapid AI innovation.

Model and Agent Development. CoreWeave provides the developer tools and integrations teams use to build, evaluate, deploy, and monitor models and agents, with experiment tracking, governance and guardrails, and workflow integrations that speed iteration from prototype to production. With Weights & Biases (acquired by CoreWeave in 2025), we provide a comprehensive, proven AI development platform and cutting-edge tools for AI development—including model pre-training (W&B Models), post-training (W&B Training), model serving (W&B Inference), and agent evaluation and monitoring (W&B Weave). For our customers, these tools mean faster, more efficient AI agent development, from training to production. Our integrated capabilities enable our customers to correlate infrastructure events with model performance while providing observability tools for debugging. In October 2025, we also acquired Marimo, creator of the marimo notebook, an open-source, AI-native environment, to unify the generative AI development workflow and enable faster, smarter AI development. And in September 2025, we acquired OpenPipe, a leading platform for training AI agents with RL. In October 2025, we launched the first publicly available serverless RL capability to build reliable AI agents. This first joint offering from Weights & Biases and OpenPipe enhances our customers' ability to train with RL—quickly, easily, and at scale. Our expanding agent development capabilities help our customers accelerate experimentation, fine-tuning, and RL at scale, seamlessly bridging model development and production deployment. To meet the specialized needs of industrial customers, in November 2025 we acquired Monolith, which provides a full-stack platform for enterprises that want to accelerate industrial innovation in the physical world through simulation and machine learning capabilities.

The Essential Cloud for AI—and for our customers

We work closely with our customers to tailor their solution based on their performance needs (e.g., GPU selection), specific use case (AI model training, agentic AI and inference, and specialized workloads), and business goals—from pioneering labs and researchers envisioning the next AI breakthroughs to enterprises seeking an AI-driven competitive advantage. In all cases, our customers benefit from CoreWeave Cloud, purpose-built for AI innovation at every layer.

Our Data Center Footprint

Our platform is powered by some of the largest and most sophisticated data centers in the world, built around cutting-edge GPU clusters and state-of-the-art network technology designed to maximize performance for AI workloads. Each component of our data center technology stack is purposefully architected to deliver highly performant networking, power, and cooling, resulting in a geographically distributed, high-density, and secure data center footprint.

We operate a distributed and interconnected portfolio of data centers across the United States, Europe, and Canada, operating in six countries total. Many of our facilities are designed to scale from tens to hundreds of megawatts of power and are connected through high-speed interconnects to support burst AI workloads. The portfolio includes a range of facility sizes, which provide us with critical flexibility. These facilities range from smaller sites that support inference workloads close to where customers need it to reduce latency, to larger sites that can support high-density training workloads where latency is less critical. Regardless of data center size, our modular basis of design provides us with the ability to scale big and fast and is built to be fungible across all AI workloads. The diversity of our infrastructure enables us to right-size and fit our services to the appropriate AI workload, from training, to inference, and to agentic workloads as requirements change over time.

8

Table of Contents

Our data centers are designed to meet the technical demands of AI computing, both today and in the future. Across the industry, data center infrastructure is trending toward higher rack density and increased power per rack, driving significantly greater cooling requirements. Our advanced closed-loop liquid cooling systems address these demands by enabling higher power density and closer rack placement, and significantly reducing water consumption. This approach enables our data centers to support and deploy the latest generation of chips at scale and for us to be among the first to bring them to market at scale—with a track record of bringing NVIDIA GPUs to market first, including GB200, GB300, and RTX Pro 6000 systems. As compute density continues to increase, these same design principles position our data centers to support future chip generations, and looking ahead, CoreWeave is expected to be among the first cloud providers to deploy the NVIDIA Rubin platform, offering customers greater flexibility and choice as AI systems scale.

To support our continued growth and scale, we are making sustained investments in both our supply chain and labor force. While the chip remains at the core of delivering AI data centers, the supporting equipment is equally critical to bringing compute to life. As a result, we have continued to evolve and expand not only our component supply chain, but also our long-lead data center equipment. We have strengthened our operations to ensure accurate planning and timely procurement of the long-lead equipment required to deliver high-performing clusters, supported by strong direct relationships with providers as well as a focus on OEM and ODM partners to meet demand at scale. In parallel, to fuel a sufficient and capable labor force, we have developed the CoreWeave Apprenticeship program, working with early-career individuals to provide hands-on training and experience across a range of specialties.

Throughout all our data centers, we combine industry-leading security and compliance with a fully managed experience, allowing customers to focus on AI innovation while we operate the infrastructure. We reinforce our infrastructure with industry leading security standards and certifications, including SOC 2 and ISO/IEC 27001, to ensure that customers are met with robust data security practices. Our security measures also extend to our physical security, where we employ rigorous standards around background checks, access control, security awareness training, and a zero-trust framework. We handle the complexity of hardware, facilities, and operations.

We continue to grow our portfolio of data centers by both expanding our existing sites and developing new sites while investing in new capabilities such as self-builds. We continue our international expansion with a focus on Europe and are evaluating launches in other regions. For more information, refer to Risks Related to Our Business and Industry included in Part I, Item 1A of this Annual Report on Form 10-K.

Go-To-Market

CoreWeave's go-to-market ("GTM") organization consists of Sales, Marketing, Field Engineering, Support, Customer Experience, Partnerships, and Revenue Operations. Our sales engine prioritizes lean and efficient scaling, leveraging AI to accelerate execution. Our organization is driven by customer-obsessed team-members focused on acquiring new customers and expanding with existing customers, who are aligned by vertical and geography. Our GTM team operates across the breadth of CoreWeave Cloud, selling and supporting products that provide foundational infrastructure for large scale research training, runtime accelerators and orchestrators, agent and AI application hosting, and model and agent development tools. For example, initiatives, such as "direct-to-expert," enable our customers to receive consultative expertise and technical support to troubleshoot issues and optimize performance. Our robust product offering enables us to diversify across industries, geographies, and customer segments, while engaging with customers throughout their journey to ensure they are positioned to develop and productionize AI.

Our Culture and Values

Our mission is to power the creation and delivery of the intelligence that drives innovation. We believe the success of this mission depends on the power of our people and a culture defined by ownership, curiosity, and trust. Our multidisciplinary team operates with this mindset and is focused on disciplined execution, running toward the hardest technical challenges, and taking ownership of solving them. We have assembled leading engineering, infrastructure, and go-to-market teams with what we believe is an unmatched knowledge base in accelerated computing at scale. These teams work synchronously to deliver The Essential Cloud for AI, purpose-built to support high-performance AI workloads for our customers.

To maintain this competitive advantage and continue to innovate, we remain focused on building an inclusive working environment in which our team members can push the boundaries of what is possible. We believe that fostering an inclusive work environment is important to maintaining our competitive advantage and ability to innovate. Our teams are composed of individuals with a wide range of skills, experiences, and perspectives, which positions us to run toward the hardest problems, address complex technical challenges, and support our customers at scale. We focus on creating an

9

Table of Contents

environment that enables employees to perform at their highest potential and contribute meaningfully to our long-term success.

We have created and instilled a set of Core Values that our employees embrace and demonstrate each day.

Our Core Values include:

•Be Curious at Your Core. We foster a mindset of continuous learning, understanding that curiosity drives innovation and is essential in staying ahead of our competition. Curiosity is a driving force that allows us to explore ideas and innovations that push boundaries, challenge the status quo, and better serve our customers.

•Act Like an Owner. We take full responsibility for our work and our decisions to find the best solution to any challenge. We take initiative, hold ourselves accountable, and foster a culture where everyone feels empowered to contribute to our success.

•Empower Employees. We trust each other to make decisions, communicate transparently, and are clear on our goals. We support each other with the tools, resources, and feedback necessary to succeed.

•Deliver Best-in-Class Customer Experiences. We go above and beyond to understand our clients' needs and aim to exceed expectations at every step. Our commitment to excellence ensures we build lasting partnerships and set the standard for exceptional service.

•Achieve More Together. Collaboration is at the heart of our success. Together, we unlock greater potential and solve challenges more effectively. We believe that by leveraging diverse perspectives, supporting each other, and working as one team, we can accomplish more than we could ever do alone.

As of December 31, 2025, we had 2,189 employees, with 1,967 in the United States and 222 in 19 other countries. We provide competitive cash and equity compensation and scale our workforce to meet the growth and operational needs of our business. Employees are equipped with the tools, systems, and resources required to operate effectively in complex, fast-moving environments. We believe our culture and the strength of our relationship with our employees distinguishes how we operate and succeed as a business.

Sustainability

AI can drive numerous benefits. However, we recognize that there are also risks that need to be managed to ensure that our platform is built to maintain resiliency and assure its longevity. One of the challenges we expect to face is to ensure that we take appropriate steps to mitigate the environmental impacts of our operations. We are focused on the benefits our CoreWeave Cloud Platform can provide while managing the associated risks.

Our strategy is centered around: Clean Energy, Efficiency by Design, Responsible Resource Usage, and Supplier Relationships.

Clean Energy

Our fleet of data centers necessarily uses a significant amount of power to enable our customers to train AI models, particularly large language and deep learning models. This energy usage results in greenhouse gas ("GHG") emissions, which impact the environment. Our goal is to minimize our environmental impact. For example, several of the existing data centers in the United States and Europe in which we are a tenant are powered by non-emitting sources of energy, and we are exploring other ways we can reduce our environmental impact.

We are currently evaluating and measuring Scopes 1 and 2 GHG emissions from our facilities and data center operations and will soon begin to evaluate our emissions resulting from our supply chain (Scope 3). A large portion of our GHG emissions are currently Scope 2 emissions, which result from purchased electricity for our data centers. The mix of our GHG emissions could shift over time as we grow and begin to control more of our data centers, potentially increasing our ability to contract larger amounts of non-emitting sources of energy.

We are tracking renewable energy sourcing to assess progress toward increasing the proportion of energy procured from low-carbon and renewable sources.

As we continue to scale, we are integrating GHG emissions management and other sustainability considerations into our lease negotiations for new data centers and the renewal of existing leases to reduce our data centers' environmental

10

Table of Contents

impacts. We also aim to establish our operations in buildings that meet high levels of environmental standards where possible.

Efficiency by Design

Our business was founded on a simple principle: AI infrastructure should be designed for efficiency from day one, not retrofitted onto legacy systems. This shapes everything we do.

We aim to build next-generation data centers in the right places, with the right partners, to ensure we can deliver performance while continuously improving efficiency.

We work with our data center partners to track Power Usage Effectiveness ("PUE"), enabling us to collectively measure how efficiently data centers use energy. We are collecting the necessary PUE data with the aim of reducing both operating costs and indirect emissions.

Responsible Resource Usage

Water Stewardship

We are committed to responsible water consumption. Monitoring water consumption helps identify opportunities to reduce freshwater use, particularly in water-stressed regions, and supports climate resilience by mitigating exposure to drought-related disruptions. CoreWeave's infrastructure incorporates innovative technologies, such as closed-loop liquid cooling systems to manage the increased heat output stemming from our use of smaller and more dense servers and improve energy efficiency.

Waste Management

In addition to managing scarce resources such as water, we intend to extend our responsible resource usage to waste management. For example, we have implemented heat recovery systems and intend to evaluate the potential to expand our ability to repurpose heat generated by our data centers, which can optimize power consumption and improve environmental performance.

Supplier Relationships

We will also look for opportunities to utilize our relationships with suppliers to mitigate the environmental impacts of their operations, while improving our data centers' environmental performance.

For more information, refer to the risk factor entitled "Investors' expectations of our performance relating to environmental, social, and governance factors may impose additional costs and expose us to new risks" included in Part I, Item 1A of this Annual Report on Form 10-K.

Competition

The AI cloud market is highly competitive and continues to evolve rapidly. We primarily compete with hyperscalers who offer general purpose cloud computing as part of a broader product portfolio, several of which are also customers of, and partners to, CoreWeave. We also compete with smaller cloud service providers.

CoreWeave's competitive advantage is a result of our platform being purpose-built for AI and accelerated compute use cases. We believe our speed to market with the latest generations of GPUs, proprietary software and orchestration platform, security standards and ability to deliver a highly attractive total cost of ownership all contribute to CoreWeave's rapid growth. Our competitive differentiation is further underpinned by our ability to service AI compute use cases of all sizes rapidly and fungibly. For more information, refer to the risk factor entitled "We face intense competition and could lose market share to our competitors, which would adversely affect our business, operating results, financial condition, and prospects" included in Part I, Item 1A of this Annual Report on Form 10-K.

Intellectual Property

Our intellectual property is an important aspect of our business and helps us to maintain our competitive position. To establish and protect our rights in our proprietary technology, we rely upon intellectual property laws and other legal protections covering patents, copyrights and trade secrets, as well as contractual restrictions such as confidentiality agreements, licenses and intellectual property assignment agreements.

11

Table of Contents

We control access to our intellectual property and confidential information through internal and external controls. We maintain a policy requiring our employees, contractors, consultants and other third parties to enter into confidentiality and proprietary rights agreements to control access to and non-disclosure of our proprietary information. Intellectual property laws and our procedures, policies, controls and restrictions provide only limited protection, and any of our intellectual property rights may be challenged, invalidated, circumvented, infringed, misappropriated or violated.

Government Regulation and Regulatory Environment

We operate in a complex and evolving regulatory landscape spanning AI, critical infrastructure, energy policy, international trade, data privacy, export controls, and national security considerations. The scope and applicability of regulations in these areas continue to broaden with new legal and regulatory developments, and their interpretation remains uncertain. We are subject to regulations across multiple jurisdictions and governmental agencies.

Like other companies in the technology and infrastructure sectors, we face heightened regulatory scrutiny from both U.S. and foreign governments. These numerous and sometimes conflicting laws and regulations may directly affect our ability to pursue our business model and offer our products and services. Our compliance with these regulations increases our cost of doing business and may require operational adjustments, changes to our business practices, or limitations on our ability to offer services in certain geographies or to certain customers.

Key areas of regulatory impact include:

AI Infrastructure and Critical Infrastructure Policy. As policymakers increasingly focus on the role of infrastructure providers in supporting national competitiveness, we engage with federal, state, local, and international authorities on emerging policies affecting AI infrastructure planning, energy supply, grid management, and workforce development. These regulations could affect our capital planning, site selection, and operational practices.

Energy and Environmental Compliance. Our data center operations are energy-intensive and subject to evolving environmental regulations, renewable energy standards, and sustainability requirements across jurisdictions where we operate. We work collaboratively with utilities and energy regulators on infrastructure scaling and grid integration. These requirements may affect our cost structure and operational flexibility.

Export Controls and International Trade. Our technology and services are subject to U.S. export controls, sanctions laws, and trade regulations governing the sale and operation of advanced computing equipment. These controls, administered by the U.S. Department of Commerce and other agencies, continue to evolve in response to national security considerations. Changes to export regulations could limit our ability to serve certain customers or markets, materially affect our competitive position in certain geographies, and require operational adjustments.

We maintain compliance programs to monitor these regulatory developments and ensure adherence with applicable laws. To date, costs and accruals incurred to comply with applicable governmental regulations have not been material to our capital expenditures and results of operations. However, certain emerging policy areas, particularly those related to AI governance, critical infrastructure planning, the cost of energy, energy infrastructure requirements, and export controls, may require material investments or operational adjustments over time. As these regulatory trends continue, our cost of doing business may increase, our ability to pursue certain business practices may be limited, and we may need to change our business operations to comply with evolving regulatory requirements. For more information, refer to Risks Related to Legal and Regulatory Matters included in Part I, Item 1A of this Annual Report on Form 10-K.

Available Information

Our website is located at www.coreweave.com. Our investor relations website is located at www.investors.coreweave.com, where users can access our Annual Reports on Form 10-K, Quarterly Reports on Form 10-Q, Current Reports on Form 8-K, and our Proxy Statements, and any amendments to these reports, free of charge, after we file or furnish them with the SEC and they are available on the SEC's website at www.sec.gov. We use our investor relations website to disclose material non-public information. Investors should monitor our investor relations website, in addition to following our press releases, SEC filings and public conference calls and webcasts. The information available at our website is not incorporated into this Annual Report on Form 10-K. In addition, we use social media, specifically our X account (@CoreWeave) and our LinkedIn page, to communicate information from time to time. It is possible that the information that we post on social media could be deemed to be material to investors.

12

Table of Contents