Amazon ends data center NDAs amid AI backlash

Amazon will stop requiring nondisclosure agreements when negotiating data center deals with local governments, following a similar decision by Microsoft earlier this year. The change, first reported by TechCrunch, comes as secrecy around AI infrastructure has fueled community resistance, with hundreds of proposed and enacted moratoriums from New York to San Francisco.
The policy shift arrives as startups race to build the next layer of AI infrastructure. Emerald AI, which raised $150 million in August for data center power technology, and Delos Data, a chip startup led by former Intel engineers that secured $100 million last month, are among those betting on the growing demand for AI-driven compute. Meanwhile, Together AI’s recent partnership with a Saudi firm to access overseas data centers suggests that startups are already looking beyond U.S. borders for capacity.
Amazon’s decision may ease some local tensions, but it doesn’t address the broader concerns driving opposition—chiefly, the strain AI data centers place on power grids and water supplies. Whether transparency alone will be enough to win over skeptical communities remains unclear. For now, the move is less about solving the problem than acknowledging it.
Sources: techcrunch.com
“Amazon’s move signals a shift toward transparency in AI infrastructure deals, but whether it eases local opposition remains an open question.”
Read the original reporting
The outlets below did the original reporting.
- Amazon drops data center NDAs, and AI agents want your credit card — techcrunch.com
Related briefs
- India’s AI startup funding jumps 265% YoY in Q3
- Juspay FY26 revenue jumps 30% to Rs 664 Cr as AI bets widen loss
- Anthropic bans abusive behavior toward AI models in policy update
- Nvidia defends open-source AI safety alliance amid scrutiny
- Edmonton startup SenseHuman AI scales AI-powered hardhat for worksites
This brief was drafted automatically from the sources above and published under our editorial policy. Spotted an error? Tell us.