Amazon Drops Data Center NDAs, but Can Transparency Rebuild Trust?

Introduction

The expansion of AI is turning data centers from largely private corporate facilities into projects with visible consequences for local communities. Amazon says it will stop using nondisclosure agreements when negotiating data center deals with local governments. Microsoft made a similar change earlier this year. The shift suggests that major cloud companies recognize a growing problem: the technical promise of AI does not automatically translate into public trust.

Data center negotiations have often taken place behind closed doors. Companies may want to protect commercial terms, while local governments weigh possible investment, jobs, tax benefits, and pressure on public infrastructure. But when residents learn about a project only after its basic direction has been set, confidentiality can look less like a business safeguard and more like a lack of democratic oversight.

Key points

• Amazon is changing its negotiating practice. The company says it will no longer use NDAs when discussing data center projects with local governments, following Microsoft’s earlier decision.
• Opposition is spreading. Resistance to AI infrastructure has contributed to hundreds of proposed and enacted moratoriums, ranging from New York to San Francisco.
• Disclosure is not the same as consent. Ending an NDA may give residents earlier access to information, but it does not by itself answer questions about energy, water, public costs, local benefits, or accountability.
• Trust is a broader AI challenge. TechCrunch’s Equity podcast also examined personal AI startups seeking access to inboxes, files, and credit cards. In both cases, adoption depends on whether people believe the organization seeking access will act responsibly.

Why it matters

The immediate value of dropping NDAs is timing. Residents, environmental groups, and public agencies may be able to examine a proposal before the approval process is effectively settled. Earlier disclosure can reduce the sense that a project has appeared without warning and can make it easier to hold local officials to their public commitments.

Still, transparency addresses only part of the trust problem. Communities typically want verifiable answers: How much electricity and water will the facility require? Who will pay for supporting infrastructure? What benefits will remain locally? Which promises will be monitored after construction? If a company simply announces that no NDA will be used, without publishing meaningful impact information or creating channels for public feedback, the change may remain largely symbolic.

The connection to consumer AI is revealing. Data center developers must persuade communities that infrastructure growth will not impose hidden costs. Personal AI companies must persuade users that highly sensitive permissions can be handled safely and within clear limits. The audiences differ, but the underlying challenge is similar: AI is expanding faster than public understanding and institutional oversight. Companies are asking for access to resources, information, or authority before trust-building systems are fully established.

Amazon’s and Microsoft’s decisions are therefore a constructive step, not a completed solution. The next standard should be continuous disclosure, public questioning, and enforceable accountability. Transparency becomes meaningful only when it gives affected people a practical way to participate, verify claims, and challenge outcomes. Without those mechanisms, removing an NDA may improve optics without changing the relationship between AI infrastructure developers and the communities hosting them.

Source: TechCrunch AI