The demonstration is part of a wider challenge to the proposed development, centred on questions of electricity, water, pollution, local infrastructure and the rapid expansion of AI-related computing. Yet some of the most widely repeated claims around the scheme go further than the published evidence currently supports. The project has received a conditional planning resolution, not a demonstrated final planning decision in the records reviewed; a construction date remains a possibility rather than a confirmed timetable; and assertions about cooling and renewable power at this specific site require more documentation.
What is proposed at Skelton Grange
Microsoft consulted locally on its Skelton Grange datacentre proposal in June 2025, ahead of submitting its planning application. The development is linked to application 25/06139/FU and is described by land developer Harworth as a hyperscale data-centre campus of approximately 500,000 square feet across three data halls and associated buildings.
Harworth also says the broader application sought outline permission for a separate warehouse or industrial-and-logistics element on neighbouring land. That distinction matters. Public discussion can flatten the scheme into one enormous “data centre,” while the planning proposal includes different elements that may have different operational and planning implications.
The scale is nonetheless substantial. A campus of this kind raises consequential questions for a city: not only the footprint of the buildings, but the network capacity, construction disruption, water strategy, drainage, noise controls, lighting, emissions safeguards and local economic case that accompany it. Those questions are legitimate whether one supports or opposes the project.
However, the available record does not settle every technical point. Conflicting accounts of the total floorspace have circulated, and the approximately 500,000-square-foot description across the three halls and auxiliary buildings is the more cautious figure to use from the material available. It would be premature to build further claims on unverified dimensions or specifications.
A planning resolution is not the same as a final approval notice
The most important procedural detail is easily lost in coverage of the protest. On 23 April 2026, Leeds’s planning panel passed a resolution to approve the Microsoft application. The decision was delegated to the Chief Planning Officer, subject to conditions and completion of a Section 106 agreement.
That represents a meaningful milestone for the developer. It is not, though, identical to evidence that every condition was discharged, the Section 106 agreement completed, and a final decision notice issued. The materials reviewed do not confirm that final stage as of 14 September 2026.
This may sound technical, but it has real consequences. A resolution to approve signals the planning panel’s intent and can set a project on a path towards construction. Conditions and legal agreements are where substantial matters can be formalised, including mitigation, contributions, transport arrangements or environmental requirements. Until the final status is demonstrable, describing the campus as simply “fully approved” risks overstating the public record.
Nor is there a confirmed near-term building schedule in the evidence available. A November 2025 industry report said construction could begin in early 2027. That was a reported possibility, not a firm start date. Harworth’s March 2026 results said enabling works were continuing and forecast completion of the Skelton Grange site in 2030. A long-range site-completion forecast should not be read as a guaranteed operational date for the data centre, but it does underline that this is a multi-year development rather than an imminent finished facility.
Protest numbers and the larger campaign
The campaign against the proposal has moved beyond online opposition. Organisers reported that they lost count after the first 300 attendees at the September demonstration. It is reasonable to report that as the organisers’ account of turnout; it is not yet an independently verified attendance total.
The campaign has also achieved a considerably larger petition response. An independent report published on 28 August said more than 41,000 people had signed a petition opposing the development. That figure is time-specific. It supports the statement that the petition had exceeded 41,000 signatures by late August, not a claim about its precise total on the day of the protest.
Neither a large petition nor a well-attended march decides a planning application by itself. But both are evidence that the proposal has become a public issue rather than a routine industrial redevelopment. For councillors, planners, regulators and Microsoft, the practical implication is clear: residents will expect detailed answers that can be checked against binding conditions and operational evidence, rather than broad assurances.
The environmental questions: real concerns, incomplete answers
Campaigners have raised concerns about power and water demand, noise, light and air pollution, foul water, waste heat and the climate impact of expanding data-centre infrastructure. These are concerns and allegations put forward by opponents; they are not measurements demonstrating that the Leeds site will cause each stated effect.
That distinction should not be used to dismiss the questions. A proposed hyperscale campus should face scrutiny over its resource use and safeguards. But it does mean claims need to be separated into three categories: what is proposed in submitted designs, what is required through planning conditions, and what will be measured when or if the site operates.
Cooling has become a particular point of contention. Campaign meeting notes say Microsoft described air-circulation cooling up to 29.4°C, with water-assisted cooling above that threshold. Those notes do not establish a final cooling system, annual water demand, source of water or expected operating pattern. Crucially, they do not substantiate the claim that air cooling will be used “around 95%” of the time.
The difference is more than semantic. The annual quantity of water used, whether it comes from mains supply or another source, how heat is rejected, and what happens during hotter conditions are central to assessing local impact. A temperature threshold discussed at a meeting is not a substitute for a published operational model. Residents seeking confidence should ask for specifications that identify peak and annual demand, the assumptions behind those forecasts, discharge and drainage arrangements, noise limits, lighting controls, and how compliance would be monitored.
The same caution applies to claims about electricity and carbon. Microsoft says it matched 100% of its annual global electricity consumption with renewable energy in its 2025 financial year and continues to target carbon-negative, water-positive and zero-waste operations by 2030. At the same time, the company reported that its total emissions increased 25% year on year, primarily due to data-centre expansion.
Those statements capture a wider tension in the technology industry: renewable-energy matching and long-term environmental targets can coexist with rising absolute emissions as physical infrastructure expands. They do not establish that the Leeds campus will have a site-specific, one-for-one renewable-electricity arrangement, nor do they answer local questions about grid capacity, water use or construction emissions.
For the public, the meaningful test is not whether a company has a global target, but what enforceable commitments apply locally. A Leeds-specific energy, water and emissions account would be more informative than importing corporate-wide claims into a project-level debate.
Why the national political response will not stop the scheme
Leeds East MP Richard Burgon is the primary sponsor of an Early Day Motion, tabled on 25 August, calling for a national moratorium on new AI data centres while environmental, public-health and infrastructure impacts are assessed.
The motion gives parliamentary visibility to the concerns voiced in Leeds and elsewhere. It does not, however, cancel the Skelton Grange proposal or automatically place the planning process on hold. An Early Day Motion is a parliamentary mechanism used to express an opinion, publicise a cause or show support for a position, and such motions are rarely debated.
That does not make the intervention irrelevant. It places the local fight within a national policy question: whether the UK’s planning, energy and environmental systems are moving fast enough to assess the cumulative effects of a data-centre buildout associated with cloud services and AI. But any actual moratorium would require a different policy or legislative route. Treating the motion as if it had already halted the Leeds scheme would mislead residents about both Parliament and planning law.
What Microsoft, planners and residents should clarify next
The available evidence points to a debate that needs greater precision, not louder but less reliable claims from either side. Microsoft and the relevant planning authorities should make it straightforward to determine the project’s final planning status, the outstanding conditions and Section 106 obligations, and the timetable for any decision notice.
They should also provide accessible, project-specific information on cooling, annual and peak water requirements, water source, drainage, noise, lighting, air-quality safeguards, waste-heat plans, grid connection and anticipated operational monitoring. If some information is commercially sensitive, that does not remove the need to disclose the environmental assumptions necessary for public scrutiny.
Campaigners, meanwhile, have a strong case for focusing on verifiable demands. The reported protest turnout and 41,000-plus petition figure show organised public concern. The most durable challenge will come from testing planning conditions and technical claims against evidence, rather than treating uncertain figures as settled fact.
For Windows and Microsoft customers, the Leeds dispute is a reminder that the future of large-scale computing is not only a matter of software features and AI investment. It is also shaped by land use, energy systems, water resilience and public consent. The dossier does not establish which services this proposed campus would run or when it might become operational. What it does show is that the project has become a concrete test of whether the benefits and burdens of new computing infrastructure are being explained with enough local detail to earn trust.
The bottom line
A substantial campaign has formed around Microsoft’s proposed Skelton Grange campus, and organisers report that more than 300 people attended its September protest. The development has cleared an important planning-panel hurdle through a conditional resolution to approve, but the reviewed records do not confirm a final decision notice or a fixed construction start.
The environmental concerns deserve serious examination, especially around water, power and local effects. Yet key claims about the Leeds facility’s cooling profile and renewable-power arrangement remain unverified. The next phase of the debate should therefore turn on documents, conditions and measurable commitments—because those, rather than slogans or corporate aspirations, will determine what the project means for Leeds.