Amazon's decision to abandon nondisclosure agreements in talks with local governments over data center projects marks a notable shift in how the biggest players in cloud and artificial intelligence are managing public scrutiny. The company said it will no longer rely on secrecy provisions when negotiating with municipalities, a practice that has long frustrated residents, local officials and advocacy groups seeking to understand the economic and environmental terms of major infrastructure deals.
The move follows a similar decision by Microsoft earlier this year and reflects a broader recognition across the sector that opacity has become a liability. As AI demand accelerates, hyperscale data centers have become the physical backbone of the industry, requiring vast tracts of land, large power commitments and heavy water and grid infrastructure. Those needs have often been negotiated behind closed doors, leaving communities to learn about projects only after plans are advanced or approvals are imminent.
Secrecy Meets Resistance
That approach has increasingly backfired. Across the United States, secrecy around data center proposals has fueled a wave of local opposition, with residents and elected officials questioning whether promised jobs and tax revenue justify the strain on electricity systems, water supplies and land use. The backlash has helped drive hundreds of proposed and enacted moratoriums from New York to San Francisco, as cities and counties attempt to slow or study the pace of development before committing to large-scale AI infrastructure.
For Amazon and its peers, the political risk is no longer theoretical. Data centers are no longer seen as neutral industrial assets; they are now tied directly to the rise of frontier AI models, which require enormous computing capacity and continuous expansion. That association has sharpened public concern, especially in places already grappling with power constraints, climate goals and affordability pressures. In that environment, NDAs have come to symbolize not just corporate confidentiality but a lack of accountability.
Trust As New Currency
The shift away from secrecy is therefore as much about trust as it is about process. By removing NDAs from local negotiations, Amazon is signaling that it understands the need to make its case in public rather than rely on private bargaining. Yet transparency alone may not be enough to defuse opposition. Communities want more than disclosure; they want enforceable commitments on energy sourcing, water use, local hiring and long-term economic benefit.
That is especially true as the AI infrastructure boom collides with a broader skepticism about whether the public is bearing the costs while technology companies capture the upside. Local governments are under pressure to attract investment and tax revenue, but they are also being asked to explain why scarce public resources should be allocated to facilities that can consume as much power as small cities. The result is a more adversarial permitting environment, one in which even well-capitalized companies must now compete for legitimacy.
The timing is also significant. A wave of startups is betting that consumers will grant AI agents access to personal data, accounts and workflows in exchange for convenience and automation. That bet depends on a baseline of trust that the industry has not yet fully earned. If users are expected to hand over sensitive information to software agents, the companies building the underlying infrastructure will also need to show they can operate with greater openness and restraint.
A Test For The Sector
Amazon's policy change does not resolve the deeper conflict over AI infrastructure, but it does acknowledge that the old playbook is failing. The industry can no longer assume that secrecy will smooth the path for expansion. In the current climate, hidden deals invite suspicion, and suspicion invites delay.
What remains to be seen is whether transparency becomes a genuine standard or merely a public-relations adjustment. If companies continue to negotiate aggressively while disclosing little about environmental impacts, grid demands or community benefits, backlash is likely to persist. But if Amazon's move is the start of a broader reset, it could help establish a new norm in which AI infrastructure is built not just faster, but more openly.
For now, the message from local communities is clear: the era of quiet data center deals is ending. The companies racing to build the physical foundation of AI will have to win not only permits and power, but public consent.