Can the AI industry persuade data center opponents by getting rid of NDAs?

Amazon has said it will no longer use nondisclosure agreements while negotiating with local governments over data centers, following Microsoft, which made a similar commitment earlier this year. Amazon's statement arrived inside a longer blog post that argued data centers are good for communities, rather than as a standalone announcement. The change was discussed on the latest episode of TechCrunch's Equity podcast by Sean O'Kane, Rebecca Bellan and Anthony Ha.
The context is a stretch of local resistance to data center construction. San Francisco recently passed a moratorium, and New York's version halts permits for large projects until the state completes an environmental review process, expected to take about a year. Bellan said the moratoriums she has heard about typically run one or two years, while the data centers at issue are not coming online for another two to five years, and buildouts face delays from zoning and funding. Her read is that the practical effect of the political pushback is uncertain. Ha pointed out that moratoriums are often framed as a pause to study the issue rather than a rejection, which leaves room for a community to later decide it supports a project.
O'Kane argued the NDA practice sits at the root of the industry's data center problem. Many projects are negotiated essentially in secret, he said, and sometimes the public knows a data center is being discussed but not which company will occupy it. He compared it to Uber's rapid expansion in the mid-2010s, when the company negotiated with local governments in secret, including a case outside Orlando involving tax subsidies for rides connected to public transit. According to O'Kane, secrecy made it easier for the company to ask for more. He welcomed Amazon's move but said the parties are entrenched and the effect is unproven.
Ha was similarly measured. He does not think dropping NDAs will turn opponents into supporters, but he sees it as addressing a basic complaint: communities do not know what is being built because negotiations happened behind closed doors. In his framing, that makes it a good first step rather than a resolution.
The panel also treated the debate as part of a broader communication problem in the tech industry. O'Kane said companies do a poor job explaining themselves and are frustrated that most people cannot picture how AI or data centers would be useful to them. He tied that to an era in which companies prefer to communicate directly and treat the press as irrelevant, and cited Tesla's "abundance" mission, renamed from "sustainable abundance," as an example: a year after announcing it with a multi-page PDF, he said the company still lacks a clear definition, and even fans criticized the document as unspecific.
Ha drew a line from that to a shift in posture. For a while, he said, the AI-driven future was presented as inevitable, with people told to get on board or be left behind. He welcomed the pushback forcing companies to make the case instead of asserting it. On Tesla, he said the striking part was not the missing definition of abundance but that the promise was made at all, which he linked to a relationship with the company that is about remaking society as much as selling cars.
For developers and product teams, the practical signal is that the physical layer of AI is becoming a political constraint, not just an engineering one. A likely trade-off is that transparency around siting, power and water use invites more organized local input and longer timelines, even if it can reduce the secrecy that fuels opposition. Inference on my part: teams planning capacity should assume permitting and community review become schedule risks, and vendors may be pressured to disclose earlier which workloads and operators a site is intended to serve. Why it matters: cloud and AI builders depend on new data center capacity, and the terms on which communities approve it are shifting. If disclosures become standard, developers may need to account for public scrutiny of energy, water and property impacts earlier in planning, while local governments gain leverage to slow or reshape projects. Whether that changes outcomes is still unclear, but the industry's willingness to negotiate openly is now part of the story.
Based on reporting from the original publisher. Visit the source for full context and later updates.
Publisher excerpt
On Equity, we debate the impact of data center moratoriums and whether the AI industry has a communication problem.