AWS Drops Data Center NDAs and Open Sources a Jev-Style AI Decision Model
aiinsider ·
Amazon Web Services has announced two significant strategic shifts, ending its use of nondisclosure agreements with government agencies for data center projects and releasing an open-source AI decision model. AWS CEO Matt Garman stated that transparency is crucial to address public backlash and environmental concerns, noting that over 100 local moratoriums on data centers are currently under consideration in the United States. Garman challenged common criticisms by asserting that direct water usage is minimal and that electricity price hikes stem from grid issues rather than data centers alone. Critics, however, argue these claims overlook broader energy impacts, citing a recent watchdog report linking data centers to significant price increases. In parallel, AWS released Strands Decider 2B, an open-source model inspired by TypeSafe’s Jev, designed for faster and cheaper agent automation. This lightweight model, built on Qwen3.5–2B, allows customers to run decisions locally with lower latency and higher reliability than full large language models. While AWS engineers believe this approach offers cost advantages, competitors suggest that building genuinely smart deciders remains technically difficult.
aiinsider