Summary

• The 84% finding concerns ministerial data centre stakeholder meetings from July 2024 to June 2026.
• The group found limited ministerial engagement with communities and environmental organisations.
• Its separate revolving-door review covers July 2024 to September 2026 and needs its own denominator.

Businesses and industry associations attended 84% of recorded UK ministerial stakeholder meetings concerning data centres between July 2024 and June 2026, according to an investigation published on 27 October by Spotlight on Corruption.

The group’s Zoned Out report argues that public engagement was poorly balanced while the government developed AI Growth Zones, which are intended to accelerate investment in computing infrastructure. The 84% statistic concerns a defined set of ministerial stakeholder meetings, not all departmental or government activity on AI.

The researchers say ministers met two community groups about data centres during the review period, both after their locations had been selected as AI Growth Zones. They found no ministerial meetings with environmental non-governmental organisations in the records they examined.

Who had access to policy discussions

Companies developing data centres can provide technical evidence about power availability, construction, investment and land requirements. That information is useful to officials designing an infrastructure strategy, but other parties may have evidence about local resource constraints and distributional effects.

Spotlight on Corruption argues that community and environmental organisations lacked comparable opportunities to influence early policy choices. The report also identifies only one meeting between ministers or the most senior officials and affected local authorities before AI Growth Zone designation.

A recorded-meeting analysis does not necessarily capture every conversation between government and outside organisations. Local planning procedures can also provide opportunities for consultation distinct from national strategy formation. Those limits should be acknowledged when interpreting the findings.

Even so, consultation that occurs only after a location has been designated may be less able to influence which areas receive priority infrastructure support. The sequencing matters because public decisions can shape land markets and network investments before an individual planning application is submitted.

Power, water and physical limits

Large AI computing facilities require substantial electrical power, while cooling, construction and network connections can create local demands. Actual resource use depends on site design, equipment utilisation, energy supply and water-management technology.

Different cooling methods have different consequences for local water use. A planned computing capacity figure cannot by itself establish future water withdrawals, and a commitment to renewable electricity does not automatically mean infrastructure imposes no additional network costs.

Assessments should distinguish developer-funded improvements from spending by regulated utilities, consumers or public authorities. Grid reinforcement and water infrastructure may involve several parties, so the distribution of costs is as important as the headline private investment.

The investigation says no national-level environmental and social impact assessment for the Growth Zone strategy had been published during its research, and it questions the role of environmental organisations in government sustainability discussions.

Those findings are attributed to the report rather than independent proof that every proposed facility fails an environmental test. Project-level planning decisions may involve separate assessments, and their quality must be evaluated on their own evidence.

Influence and movements into industry

Spotlight on Corruption separately identified 38 movements from ministerial and senior civil service roles into AI companies between July 2024 and September 2026. This is a different period from the 84% meetings analysis, which ends in June.

Taking employment in the technology sector after public service is not, by itself, evidence of misconduct. Assessing a particular appointment would require details of former duties, commercial interests, applicable restrictions and the approvals obtained.

The report considers that such movements can heighten questions about perceived influence when combined with uneven consultation. Its interpretation should be distinguished from the underlying count of jobs or appointments.

Competing policy considerations

The government has defended its infrastructure approach, pointing to expected investment, jobs and commitments on clean electricity and responsible water use. Those statements form part of the policy response, although they do not directly answer each reported finding about access to ministerial meetings.

A separate industry taskforce recently proposed changes to electricity connections and community benefits, as Techopia reported. That proposal is distinct from the new investigation into how the Growth Zone programme was shaped.

Spotlight on Corruption recommends a pause in new Growth Zone development pending stronger rules, public reporting of engagements and environmental assessment. These are advocacy proposals; no such moratorium has been established by the report itself.

Delaying major projects could affect investment and power-system planning, while proceeding without adequate assessment might leave costs or impacts insufficiently understood. A credible response requires evidence about both economic benefit and local consequences rather than assuming either outcome.

The Zoned Out report creates a new public record of stakeholder engagement. Further scrutiny will depend on how government explains the selection process, the role of consultation and the evidence underpinning infrastructure commitments.

Data centre developments can produce different kinds of employment during construction and operation. Evaluations should separate temporary building activity from continuing roles and distinguish jobs generated locally from positions moved from elsewhere.

Large electricity customers may also affect when network upgrades are undertaken. The existence of a connection agreement does not determine who ultimately pays each cost, making regulatory and contractual details relevant to public discussion.

Public registers of meetings and published environmental assumptions could help compare commitments with subsequent delivery. Such transparency need not disclose commercially sensitive technical information to establish the parties consulted and the broad evidence considered.