Complaint accuses EU of hiding data centres’ energy and water use
THE SO WHAT
If per-facility power and water data becomes disclosable under the Aarhus Convention, hyperscale buildout in Europe turns into a public permitting and PR exercise, not just a zoning one. Data center operators should assume site-level resource intensity will be scrutinized by NGOs and local governments and start instrumenting, auditing, and packaging that data now.
READ THE SOURCE
MORE FROM THE WIRE
Deep & Emerging TechFedEx is buying 2,000 electric trucks for $300M, and investing in the company building them
FedEx putting $300M into 2,000 EV trucks — while Europe scrambles with 10–20x depot power needs and failed peers — shows fleet decarbonization is now a grid and financing problem, not a vehicle problem. Logistics and retail operators need to model power availability and local grid upgrades into any EV or autonomy roadmap, not just TCO on the trucks.
Deep & Emerging TechMassive Pentagon hack sees records of 2.7 million US military personnel leaked during months-long data breach — names, military service records and Social Security numbers all revealed
A months-long breach exposing 2.7M military records via a file-sharing bug underlines that identity and HR data stores are now nation-state-grade targets. Treat any system holding employee PII or access metadata as critical infrastructure — lock down third-party file tools, tighten audit trails, and assume compromise when designing internal AI and data access.
Deep & Emerging TechVolantis, which aims to use vertical-cavity surface-emitting lasers, used in the iPhone's Face ID, to transmit data between AI and memory chips, raised $88M
Volantis raising $88M to use VCSELs for AI–memory interconnect is a bet that bandwidth and energy between chips — not just FLOPs — is the next bottleneck. If you’re planning large-scale training or inference, start tracking memory and interconnect roadmaps as closely as GPU counts when modeling future cost and performance.
Deep & Emerging TechCarlyle Warns on AI Hype and Credit Risk Ahead
Carlyle flagging that private credit is crowding into data centers and hyperscalers — echoing the ‘SaaS apocalypse’ pattern — is a warning that AI infra could be over-levered before demand fully materializes. If you’re building in this stack, secure long-term capacity and financing now, but avoid assuming today’s cheap credit and valuations will persist through the next rate or sentiment cycle.