NASDAQ: IREN

We also have Bitcoin mining operations. During the year ended June 30, 2026, we commenced decommissioning Bitcoin mining hardware and reallocating power and data center capacity toward AI Cloud Services, and aim to substantially complete the transition by December 31, 2026.

Data Centers

We develop, own and operate our data centers, including the associated land, grid connections and substations. This data center ownership model gives us direct control over site selection, design, procurement, construction, commissioning and ongoing operations. We believe that this control allows our business to benefit from more sustainable cash flows and operational flexibility relative to operators that rely upon third-party colocation services or short-term land leases, each of which may be subject to, among other things, termination rights, profit sharing arrangements and/or potential changes to contractual terms, such as pricing.

Our data centers are purpose-built for power-dense computing, with each element of the architecture, including power, cooling and networking, designed to support high-performance GPUs, Central Processing Units (“CPU”) and storage at scale. We primarily use modular designs that can be adapted for different compute architectures and customer requirements. Our data centers utilize a range of highly efficient cooling technologies, including direct liquid-to-chip cooling, free-air cooling, and free-air cooling supplemented with chilled coil cooling to optimize for site-specific environmental conditions where necessary. Our liquid cooled data centers incorporate closed-loop cooling systems, which support the rack densities and thermal requirements of next-generation AI compute while minimizing water consumption. We continue to refine our designs to improve energy efficiency, operating reliability, deployment flexibility, maintainability and deployment speed.

We focus on securing grid-connected power access for our data centers. Unlike behind-the-meter arrangements, which generally depend on one or more dedicated power generation sources located at or near the site, grid-connected facilities can access electricity from a broader and more diversified pool of generation resources through the transmission system. We believe this supports a more reliable, flexible and scalable long-term supply of power, including by reducing dependence on the availability and operating performance of any single generation source.

We generally target development of data centers in regions where there are low-cost and attractive renewable energy sources. Each of our operating sites has been 100% powered by renewable energy (whether from clean or renewable energy sources or through the purchase of RECs) since commencement of operations. We procure electricity under arrangements tailored to the structure of the applicable regional power market. For example, in British Columbia, where the electricity market is regulated, BC Hydro operates the transmission system and supplies electricity to our facilities under regulated tariffs. In Oklahoma, where the electricity market is also regulated, the utility through which we operate is the Public Service Company of Oklahoma. In Texas, ERCOT operates the electricity grid and administers the competitive wholesale market, and we procure electricity under market-based arrangements linked to wholesale energy prices. From time to time, we may enter into electricity derivatives or other hedging arrangements to fix a portion of our electricity costs for specified periods and reduce our exposure to wholesale price volatility. As our AI Cloud Services operations expand, we are also evaluating longer-term power purchase agreements and other structures that may provide greater price certainty and better align the duration of our electricity supply arrangements with our long-term data center investments and customer commitments.

Our data center footprint covers North America, Europe and Asia Pacific. Geographic diversification helps reduce our exposure to a single jurisdiction, transmission network, regulatory regime, energy market, climate or weather pattern, and positions us to serve customers in different regions with different data residency, sovereignty and sustainability requirements.

As of June 30, 2026 our announced data center projects and their total planned power capacity (gross MW) were as follows:

Location Capacity (MW)

Canal Flats, British Columbia, Canada 30MW

Prince George, British Columbia, Canada 50MW

Mackenzie, British Columbia, Canada 80MW

Childress, Texas, United States 750MW

Sweetwater 1, Texas, United States 1,400MW

Sweetwater 2, Texas, United States 600MW

Kiowa, Oklahoma, United States 1,600MW

Bundey, South Australia, Australia 800MW

Badajoz, Extremadura, Spain 300MW

Compute

The compute layer of our platform includes GPUs, CPUs, servers, storage and high-speed networking configured for large-scale AI training and inference. As of June 30, 2026, our GPUs installed and on order included AMD MI350X and NVIDIA H100, H200, B200, B300, GB300 and VR200 systems. Our multi-generation compute portfolio enables us to match infrastructure to different customer workload, performance and cost requirements.

Depending on customer requirements, we offer compute through dedicated bare metal and managed cloud environments. Our GPU clusters can be configured with NVIDIA InfiniBand NDR and XDR interconnects to provide high-bandwidth, low-latency communication between GPUs and servers, supporting distributed AI training and inference across large-scale clusters. We also offer a range of high performance storage solutions to address differing customer data-access, throughput and scalability requirements.

We maintain relationships with semiconductor manufacturers, original equipment manufacturers (“OEMs”), original design manufacturers (“ODMs”), and other infrastructure suppliers, including NVIDIA, AMD, Dell Technologies, Lenovo, Supermicro, Gigabyte and leading storage and networking providers. We work with these partners to procure and deploy servers and racks, high-speed interconnects, storage systems and other supporting infrastructure, and to prepare our facilities for successive generations of compute architecture. We believe our procurement scale, deployment experience

and direct control over the data center layer enables us to coordinate long-lead equipment, integrate complex infrastructure and bring new generations of compute into service rapidly and at scale.

We are an NVIDIA Cloud Partner and secured NVIDIA Preferred Partner status during fiscal year 2026. We have also achieved NVIDIA Exemplar Cloud status for our NVIDIA HGX B300 and GB300 NVL72 platforms, with the GB300 NVL72 designation awarded following NVIDIA’s testing of our NVIDIA GB300 NVL72 deployments at our Horizon 1 data center in Childress in August 2026. Exemplar Cloud providers have collaborated with NVIDIA to optimize their infrastructure to help ensure their customers’ workloads will be performant, secure, and reliable, based on real-world workload and total cost of ownership metrics.

During fiscal year 2026, we entered into a strategic partnership with NVIDIA to support the deployment over time of up to 5GW of NVIDIA DSX-aligned AI infrastructure across our global data center pipeline. Through this partnership, we intend to collaborate with NVIDIA on the deployment of NVIDIA accelerated compute in DSX AI factories, combining NVIDIA’s AI systems and architecture with our capabilities across data centers, compute and software.

Software

The software layer of our platform represents the managed services and enterprise support that enable customers to provision, deploy, manage and monitor AI workloads. We believe these capabilities expand the range of customers and use cases our platform can serve, and build upon the value of the data center and compute layers beneath them.

On August 4, 2026, we completed the acquisition of Mirantis, a provider of cloud software and services with a track record of serving more than 1,500 enterprise customers globally. Mirantis is an inaugural partner of the NVIDIA AI Cloud Ready Initiative and has integrated its k0rdent AI platform with NVIDIA DSX OS software components. k0rdent AI is a software infrastructure platform built on open-source technology, designed to manage and optimize AI infrastructure at scale across distributed environments. Its capabilities include GPU provisioning and lifecycle management, template-based deployment of clusters and services, workload orchestration and scheduling, centralized monitoring and logging, cost and consumption visibility, infrastructure upgrades, and multi-cluster operations. Together these enable organizations to improve utilization, support metering and monetization, and build differentiated AI Cloud Services.

We believe the acquisition of Mirantis and integration of k0rdent AI together strengthen three areas of our AI Cloud Services offering: deployment capability, including faster and more repeatable provisioning of customer environments; operational visibility, including monitoring and performance management; and customer support, including enterprise technical support and service delivery.

Bitcoin mining

We continue to operate Bitcoin miners at certain of our data centers while we transition this data center capacity toward AI Cloud Services. We aim to substantially complete this transition by December 31, 2026. Bitcoin mining revenue is generated by contributing computing power, or hashrate, to the Bitcoin network and receiving a share of block rewards and transaction fees.

As of June 30, 2026, our installed Bitcoin mining capacity was approximately 23.2 EH/s, representing approximately 380MW of data center capacity. Mining performance is principally affected by our share of global network hashrate, network difficulty, Bitcoin price, transaction fees, miner efficiency and availability, power cost and data center operating performance. We monitor these factors and operate or decommission miners based on expected returns and the requirements of our AI Cloud Services expansion program.

We typically liquidate the Bitcoin we mine on a daily basis and convert the proceeds into fiat currency to fund operating and capital expenditures. We held no Bitcoin on our balance sheet as of June 30, 2026.

Our Strategy and Growth Opportunities

Customers

We target AI Cloud Services customers across several segments, each with different capacity, service and support requirements:

•Hyperscalers. Large technology companies that require significant, dedicated compute capacity, often on a bare metal basis.

•Enterprises. Organizations deploying AI into products and business processes that may require managed services, enterprise support and flexible capacity.

•AI developers and frontier labs. Companies developing and operating foundation models and AI applications that require scalable compute and may use either bare metal or managed services.

•Channel partners. AI cloud providers, platforms and other intermediaries that resell IREN capacity through white-label or integrated services.

Our strategic priority is to broaden and diversify our customer base over time across customer segments, industries, geographies and workload types.

We primarily offer our AI Cloud Services under multi-year reserved capacity arrangements, which we expect to complement over time with on-demand service offerings. Reserved or committed contracts generally specify the amount and type of capacity, service levels, pricing, contract term, customer prepayments, deployment schedules, testing and acceptance conditions and ramp periods. On-demand arrangements generally provide customers with shorter-duration or usage-based access and may result in more variable utilization and pricing. Revenue generally begins only after the applicable compute has been delivered, commissioned, placed in service and accepted by the customer (where applicable).

Our approach to customer selection is focused on building durable, long-term relationships with a diverse range of counterparties whose requirements align with our platform and deployment plans. In evaluating customer opportunities, we consider factors including credit quality, strategic fit, contract duration and structure, expected utilization, pricing, prepayments and other credit support, financing implications and overall risk-adjusted returns. We seek to structure contracts that support efficient financing of the associated infrastructure, provide appropriate revenue visibility and downside protection, and generate attractive returns on invested capital over the life of the deployment.

We have made, and expect to continue making, significant investments in our sales and marketing capabilities to expand our customer base and increase awareness of the IREN platform. We develop our customer pipeline through direct engagement with prospective and existing customers, strategic technology relationships, channel and partner referrals, industry events and targeted marketing activities. We intend to focus our go-to-market investments on customer segments and industry verticals that are experiencing increasing AI adoption and have the potential to generate long-term demand for AI Cloud Services.

As we scale, we are also investing in building a global brand consistent with the scale of our growth ambitions. These investments include sponsorships, industry events, targeted marketing campaigns and other brand-building activities intended to increase awareness, support customer acquisition and strengthen relationships with technology and commercial partners. Our brand and stakeholder engagement activities also support our relationships with governments and communities in the regions where we operate or pursue development, including by communicating the economic, employment, innovation and community benefits associated with our investments.

Sustainability

Our sustainability strategy focuses on energy sourcing, efficiency by design, responsible resource use and community benefit. Our current data center operations in British Columbia are connected to the BC Hydro network and have been 100% powered by renewable energy since commencement of operations (currently approximately 98% of electricity used is sourced from clean or renewable sources, including through hydroelectricity facilities and other sources like wind, solar and biomass, as reported by BC Hydro and the remaining approximately 2% is accounted for by the purchase of RECs). Furthermore, our Childress site is located in the Panhandle region of Texas, which generates significant renewable energy. We purchased RECs in respect of 100% of our energy consumption through to June 30, 2026 at our Childress site.

Our designs are intended to reduce water consumption compared with conventional approaches. We also consider land use, noise, waste, equipment lifecycle, grid impacts and community priorities in the design and operation of our facilities.

Security

Security, compliance and operational resilience are embedded across our AI Cloud Services platform and are designed to protect customer workloads and data through controls including identity and access management, workload isolation, encryption, audit logging, vulnerability management, incident response, personnel screening, controlled facility access and security awareness training. Our Security and Compliance team maintains SOC 2 reports covering certain parts of our organization and is working toward organization-wide SOC 2 coverage and ISO/IEC 27001 certification.

Our Strengths

Experienced leadership and deep technical expertise across the AI infrastructure stack

Our Board, management team and workforce bring experience financing, developing, building and operating large-scale infrastructure, renewable energy projects, data centers and managed technology services across North America, Europe and Asia Pacific. We recently strengthened our executive team through the appointments of a Chief Product Officer, Chief Development Officer, Chief Information and Security Officer and Chief Marketing Officer, to add dedicated leadership across product, development, cybersecurity, procurement, data center delivery and marketing.

Strong AI ecosystem relationships

We maintain relationships with leading semiconductor manufacturers, OEMs, ODMs and infrastructure suppliers, including NVIDIA, AMD, Dell Technologies, Lenovo and others. In May 2026, we entered into a strategic partnership with NVIDIA intended to support the deployment of up to 5GW of NVIDIA DSX-aligned AI infrastructure across our global pipeline over time. We also granted NVIDIA the right to invest up to $2.1 billion in Ordinary shares, subject to delivery of up to 600,000 GPUs, and certain regulatory conditions. These relationships support product access, engineering coordination, procurement and deployment of new compute generations.

Strong community partnerships and support for our operations

Establishing and maintaining community support for our operations is a core part of our development strategy. Data center projects can raise community concerns regarding power availability and cost, water consumption, noise, traffic, land use, visual impact, environmental effects and pressure on housing and local infrastructure. We seek to engage early with utilities, local governments, landowners, Indigenous and First Nations communities, workforce and educational institutions and other stakeholders; select appropriately zoned sites or sites that may be capable of rezoning; design for efficient use of power and water; and communicate the expected economic and community benefits of each project.

We believe that our communities in British Columbia and Texas have benefited from jobs, local procurement, tax revenue, scholarships, community grants and workforce programs. Recent initiatives have included the IREN Scholarships & Bursaries program, community grant programs in Prince George, Mackenzie, Sweetwater and Childress, and the IREN Ignite paid summer employment program. As our footprint expands, we intend to develop locally appropriate engagement and benefit programs in Oklahoma, Spain and Australia, including workforce development, local partnerships and support for education and community priorities. We believe our community strategy supports more durable development outcomes.

Diversified access to capital

Large-scale AI infrastructure is capital-intensive, and access to cost-effective capital is a key constraint on industry growth. We believe we have demonstrated an ability to raise capital across public and private markets, providing us with the flexibility to finance data center development and deploy compute at scale. To date, our primary sources of financing have included customer prepayments, common equity, convertible notes, subsidiary-level equipment financing, including approximately $3.6 billion of investment-grade rated GPU financing in May 2026. This diversified funding platform reduces reliance on any single source of capital and helps us to align financing structures with customer contracts and the expected lives of the underlying assets.

Our Competition

The AI Cloud Services market is highly competitive and rapidly evolving. As a global AI Cloud Services provider, we compete with hyperscalers, including Amazon Web Services, Google Cloud, Microsoft Azure and Oracle Cloud, as well as specialized AI Cloud Services providers, including CoreWeave, Nebius, Crusoe, Lambda, Nscale, SpaceX and others. Certain companies may be our customers, suppliers or partners in one context and our competitors in another.

Competition also occurs at individual layers of the AI infrastructure stack. We compete with data center developers and infrastructure owners for land, power, interconnection capacity, equipment, labor and capital; with other cloud providers for GPUs and customers; and with software providers offering orchestration, monitoring and AI infrastructure management tools.

We believe the principal competitive factors in our industry include access to secured and energized power; ability to procure and deploy current and next-generation compute; speed and scale of construction and commissioning; access to capital; price and total cost of ownership; workload performance and reliability; networking, storage and software

capabilities; security and compliance; customer support; geographic coverage; and customer and technology partner relationships. Many of our competitors have greater financial, technical or commercial resources, longer cloud operating histories or larger customer bases, compared to us.

Employees and Human Capital Resources

As of June 30, 2026, we had 685 employees globally, reflecting significant workforce expansion during fiscal year 2026, including more than 60 employees added through the Nostrum Group acquisition. In addition, in August 2026 we added approximately 580 personnel through the Mirantis acquisition.

We also engage part-time employees, temporary employees, contractors and consultants as necessary to support our operations. For example as of June 30, 2026 approximately 3,000 personnel were engaged across our sites in the United States, including employees of contractors and subcontractors. None of our employees are represented by labor unions.

Our team has expertise across the full AI infrastructure stack, from power and data center development through compute deployment, networking, storage, software orchestration, managed services and enterprise support. We believe these capabilities enable us to design, build and operate integrated AI infrastructure at scale, reducing coordination risk and accelerating execution.

We believe that an engaged and inclusive culture is important for the success of our business, and we consider our employees to be the foundation for our growth and success. As such, our future success depends in large part on our ability to attract, train, retain and motivate qualified personnel. The growth and development of our workforce is an integral part of our success. We also strive to develop and foster a culture of collaboration that includes a broad range of backgrounds and perspectives and know that a company’s ultimate success is directly linked to its ability to identify and hire talented individuals from all backgrounds and perspectives.

We believe that diversity of thought is a key factor to achieving innovation and success in our industry. We seek to foster a culture of inclusivity, where diverse perspectives and experiences thrive. We endeavour to increase such diversity within our workforce and create an environment where everybody is empowered to excel.

To demonstrate our efforts, we:

•leverage inclusive recruitment practices that attract talent from a broad range of backgrounds and perspectives;

•invest in the professional growth of our employees, promoting access to learning and career development opportunities; and

•seek to actively engage with the communities where we operate, and support initiatives that promote inclusivity and education including partnering with schools and training authorities to develop training programs for the local workforce.

Government Regulation

We monitor developments in government regulation and maintain compliance programs designed to address applicable requirements. Compliance may increase our costs, require changes to our products or operations, delay projects or limit our ability to serve certain customers or markets. We are unable to predict the effect that any future regulatory change, or any overlapping or unclear regulations, may have on us, but such change, overlap or lack of clarity could be substantial and make it difficult for us to operate our business or materially impact the market for digital assets that we mine or may mine in the future.

AI Cloud Services and Data Center Regulations

We operate in a complex and evolving regulatory environment spanning AI Cloud Services, critical infrastructure, energy, environmental, health and safety, data privacy, cybersecurity, international trade, export controls and national security. Laws and policy initiatives in these areas may affect our site selection, capital planning, customer eligibility, hardware procurement, software design, data handling and ability to offer services in particular jurisdictions.

Our AI Cloud Services may be subject to laws and regulations governing privacy and data protection, cybersecurity, cross-border data transfers, critical infrastructure, online services and the development or deployment of AI systems. Requirements may differ by jurisdiction and may impose obligations relating to security controls, incident reporting, risk management, transparency, recordkeeping, data localization, customer contracting or oversight of certain AI uses.

Advanced computing equipment, software and technical services may also be subject to U.S. and other export controls, sanctions and trade restrictions. Changes to these rules could affect the GPUs and other technology we may procure, the customers or end uses we may support, and the countries in which we may deploy or provide access to capacity. Data center development and operation are also subject to permitting, zoning, utility, grid, environmental and workplace requirements, which continue to evolve as policymakers assess the power, water and community impacts of AI infrastructure.

U.S. Regulations

Government regulation of large-scale data center operations continues to evolve in the United States. Multiple federal and state regulators have expressed interest in oversight of data center infrastructure, AI Cloud Services, and the energy-intensive computing activities associated with large-scale data center operations. Federal agencies have increased scrutiny of energy use by large-scale data center operators. Future regulatory action related to the energy usage of data center operations, including possible reporting or operational requirements, could affect our business.

In July 2025, the President signed an executive order titled “Accelerating Federal Permitting of Data Center Infrastructure,” which directs federal agencies to streamline permitting and environmental review for large-scale AI data center projects requiring more than 100MW of new electrical load. The executive order also directs the Departments of the Interior, Energy, and War to authorize data center construction on appropriate federal lands, and instructs the Secretary of Commerce to launch an initiative to provide financial support for qualifying projects. The scope and durability of these measures remain uncertain and subject to change.

State level regulation of large-scale data center operations is also developing. In June 2025, Texas enacted Senate Bill 6 (“SB 6”), which introduces significant new requirements for large-load electricity customers, including data centers within the ERCOT region. SB 6 applies to customers with loads exceeding 75MW at a single site and requires such customers to contribute to transmission interconnection costs, disclose duplicative interconnection requests, and maintain backup generation or load curtailment capability during grid emergencies. Facilities interconnecting after December 31, 2025 must install remote-disconnect equipment to enable ERCOT-directed load shedding during grid stress events. The Public Utility Commission of Texas is required to review and potentially revise the methodology for allocating wholesale transmission charges to large-load customers by December 31, 2026.

On August 3, 2026, Governor Greg Abbott of Texas issued a directive to the Public Utility Commission of Texas and ERCOT to conduct a comprehensive verification and audit of all data centers advancing through ERCOT’s interconnection process before any additional data centers are approved to move forward. On that same day, ERCOT issued a market notice regarding Batch Zero timelines and processes. By August 7, 2026, ERCOT pursuant to Batch Zero as approved by the Public Utility Commission of Texas, was to notify each interconnecting distribution service provider and transmission service provider of how any Large Load was classified in the forthcoming Batch Zero Interconnection Study. These classifications were to be: load that has already been sufficiently studied for interconnection and therefore is considered base load for the Batch Zero Interconnection Study; load that requires additional study in Batch Zero and would be considered studied load in the Batch Zero Interconnection Study; and load that has not met sufficient criteria to be included in Batch Zero and therefore will require study in a future interconnection process. ERCOT’s August 3, 2026 market notice stated that it was not going to be notifying each interconnecting distribution service provider and transmission service provider of how any Large Load was to be classified. On August 10, 2026, ERCOT filed with the Public Utility Commission of Texas its “Requests for Good Cause Exceptions Relating to Batch Zero Deadlines and Status Update on Additional Matters Including the Long-Term Load Forecast.” In these requests, ERCOT stated that it is currently developing a comprehensive process to verify that all large loads included in Batch Zero satisfy ERCOT’s planning guide and to collect additional information from developers of data centers and virtual currency mining facilities including the community impact information described in Governor Abbott’s August 3rd letter. According to ERCOT, this process is expected to take several months. We expect, based upon ERCOT’s August 10th requests to the Public Utility Commission of Texas, that once this process is completed ERCOT will then classify as described above the large loads submitted as part of the Batch Zero process and begin the Batch Zero Interconnection Study. On August 20, 2026, the Public Utility Commission of Texas issued an order granting ERCOT’s requests for good cause exceptions providing the time requested by ERCOT to develop and implement the process to verify all large loads included in Batch Zero to satisfy ERCOT’s planning guide and to collect the community impact information sought by Governor Abbott. Due to recent changes in ERCOT’s Batch Zero procedures, there may be delays in the energization of, or changes to the energization levels at, projects in Texas, including the Company's new and existing projects.

Oklahoma’s Data Center Customer Ratepayer Protection Act of 2026 (the “OK Ratepayer Protection Act”), effective July 1, 2026, protects existing residential, commercial and industrial customers from paying unjust rates resulting directly

from electric service to large-scale energy users, including new data centers, new cryptocurrency mining operations and new AI computing facilities, that contract to add 75MW or greater electric load per facility or in aggregate behind a single point of interconnection to an electric supplier’s load after the effective date (each, a “large load customer”). Large load customers do not include residential, commercial agricultural or industrial ratepayers, or those entities that build generation for behind-the-meter projects. The OK Ratepayer Protection Act requires the Oklahoma Corporation Commission to ensure that all rates are fair, just and reasonable, and that costs and revenues are assigned and allocated among customers in accordance with cost causation principles. The OK Ratepayer Protection Act requires electric suppliers to establish and maintain separate terms and conditions for large load customers that (i) establish separate tariffs for large load customers, (ii) include credit requirements and any other measures necessary to ensure that large load customers reimburse the electric supplier for all costs fairly allocated to them, and (iii) for the term of service for large load customers to be at least 10 years. The utility serving our Kiowa project has submitted a proposed rate for this rate class, which shall be subject to review by an administrative law judge and a subsequent vote of the Oklahoma Corporation Commission. Additionally, the OK Ratepayer Protection Act requires any large load customer developer to notify adjoining landowners, county commissioners, and the Oklahoma Corporation Commission within 60 days of acquiring land for a qualifying project.

Additionally, Oklahoma’s SB 259, which is slated to become effective on November 1, 2026, prohibits data centers from using groundwater in open-air evaporative cooling systems or other cooling technology that consumes groundwater through evaporation or discharge without recirculation. Further, to receive a groundwater permit, a data center must demonstrate that it will use low-consumptive cooling technology.

These and other federal and state level developments may increase our compliance obligations, affect economic terms for power, or restrict siting or loading of our AI Cloud Services operations. Regulatory frameworks include environmental requirements, zoning and land use considerations, cybersecurity expectations, and requirements related to data privacy. These obligations may increase over time as governments respond to growth in data center activity and increasing demand for power. The effect of future regulatory changes at the federal or state level is difficult to predict. Any such changes could materially affect our operations, energy costs, customer demand, or the profitability of our business.

Regulation Outside the U.S.

Similarly to the United States, regulation of large-scale data center operations continues to evolve quickly elsewhere globally, and various regulators have expressed interest in oversight of data center infrastructure, AI Cloud Services, and the energy-intensive computing activities associated with large-scale data center operations. Future regulatory action related to the energy usage of data center operations, including possible reporting or operational requirements, could affect our business.

In August 2025, the Spanish Ministry for Ecological Transition and Demographic Challenge submitted a Draft Royal Decree regulating the energy efficiency and sustainability of data centers for public consultation and hearing process. This Draft Royal Decree partially incorporates and develops certain European regulations in Spain, while introducing additional obligations beyond what is required by these regulations. The Draft Royal Decree contemplates (i) an annual obligation to report environmental and socio-economic indicators for data centers with a total rated energy input above 1MW, (ii) an obligation to reuse residual heat unless a facility-level cost-benefit analysis demonstrates that this is not technically or economically viable, (iii) an obligation for data centers with an installed IT power demand of 1MW or more to report on how they take into account the best practices set out in the European Code of Conduct on Data Centre Energy Efficiency (a voluntary initiative aimed at data center operators, owners, and stakeholders, encouraging them to reduce energy consumption cost-effectively while maintaining mission-critical operations), (iv) and for data centers with a power capacity of over 100MW, the requirement to provide evidence that they are among the top 15% of facilities with the best sustainability indicators. As currently drafted, compliance with these obligations would be a condition to the grant and continued validity of the electricity grid access and connection permits required to operate a data center in Spain. The Draft Royal Decree remains in the pre-legislative phase and its final scope and requirements may change materially before enactment.

In Spain, this regulatory approach has since been reinforced at a statutory level. Royal Decree-Law 7/2026, of 20 March 2026, in its first additional provision, anticipates a further royal decree that will establish additional sustainability requirements for data centers connecting to the Spanish transmission and distribution networks, with non-compliance expressly identified as a ground for loss of grid access and connection permits or penalties. As its preamble makes clear, the aim is to ensure that the rapid growth of data center demand on the Spanish grid is matched by new renewable generation. Together with the Draft Royal Decree, the grant and continued validity of grid access and connection permits for Spanish data centers will increasingly depend on ongoing compliance with a broadening set of sustainability obligations, whose final scope will only be known once the relevant royal decrees are adopted.

Other European countries are pursuing regulatory frameworks. For example, previously, Germany transposed the EU Energy Efficiency Directive (which sets the European Union’s energy efficiency targets and establishes ‘energy efficiency first’ as a fundamental principle of European Union energy policy) into national law in 2024, mandating 100% renewable power use for data centers by 2027. That deadline would be deferred to January 1, 2030 under a draft amendment approved by the German government but not yet adopted by its Parliament.

British Columbia recently introduced legislative amendments that intend to restrict the electrical capacity available for new data center projects. As of February 1, 2026, the allocation of new electrical capacity for data center purposes in British Columbia is subject to aggregate limits that are allocated under a competitive process administered by BC Hydro.

We are subject to an evolving regulatory landscape for data centers and AI-related infrastructure in Australia. Whilst there is currently no single Australian statute specifically regulating data centers or AI Cloud Services, various Australian federal, state and territory governmental bodies are considering, or have proposed, measures targeting energy-intensive computing infrastructure, including data centers. For example, the Energy and Climate Change Ministerial Council (“ECMC”) has asked the Australian Energy Market Commission (“AEMC”) to provide targeted advice on regulatory pathways to require data centers to fully offset their demand by investing in renewable generation and firming, and by providing demand flexibility. Pursuant to this, the Australian Government Minister for Climate Change and Energy, the Hon. Chris Bowen MP, has lodged requests with the AEMC to amend the National Electricity Rules so that data center operators would pay for the network costs that they cause or accelerate, together with any associated network infrastructure costs. The ECMC has separately flagged its intention to require data centers to fully offset their electricity demand by investing in renewable energy generation and demonstrating firmed capacity. Each Australian state and territory is separately considering how to regulate data centers, including how to address energy, water, and renewable energy policy concerns associated with data center growth. If implemented, these or similar measures could increase our network, connection or compliance costs in Australia, affect the economics or timing of our development activities in Australia, or otherwise restrict our ability to operate or expand data center capacity in Australia in the manner we currently intend.

The AI market in Australia continues to evolve at a rapid pace, with growing demand from customers and the development of new technologies. These features, together with regulatory change, including in the areas of privacy, data governance and intellectual property, may impact our customers in the future and, consequently, demand for our AI Cloud Services.

Data Privacy and Security Laws

Numerous laws, regulations and standards govern the collection, use, access to, confidentiality and security of personal information (such as health or financial information) and other types of regulated information (such as health or financial information), data breach notification requirements and critical infrastructure requirements, and could apply now or in the future to our operations or the operations of our partners. Such privacy and security laws, regulations and other related obligations are constantly evolving, may conflict with each other to complicate compliance efforts, and can result in investigations, proceedings, or actions that lead to significant civil and/or criminal penalties and restrictions on data processing.

In the U.S., numerous federal and state laws and regulations, including data breach notification laws and security laws and consumer protection laws and regulations govern the collection, use, disclosure and protection of personal information. In Canada, existing federal, provincial and territorial laws and regulations govern privacy, data protection, cybersecurity, consumer protection, anti-spam, data breach notification. Some provincial laws already regulate automated decision-making technology; in addition there is proposed legislation at the federal level that will also regulate automated decision-making data processing, impose new privacy and data protection obligations, create fines and penalties commensurate with those under the General Data Protection Regulation and establish a new Digital Safety Commissioner to regulate certain activity on the Internet.

Refer to “