Claim report cards
Five common claims, checked against the evidence
Each report card states the claim, assigns an attribution grade, explains the evidence and identifies the measurement needed for a narrower conclusion.
recurring claims have documented AI contribution in the cited evidence.
βAI is driving the return of U.S. electricity-demand growth.β
Supported at the national demand levelYes, at the national level. Public data measures total data-center demand, and server models attribute a large share of the growth to AI systems.
Open report card βContestedβAI raised my household electricity bill.β
The effect is utility-specificA data center changes a utilityβs costs. Its rate agreement, dedicated-upgrade charges, fuel prices and cost allocation determine the effect on household bills.
Open report card βScapegoatedβAI caused the transformer shortage.β
Overstated: AI adds demand to an older shortage with several causesDOE traces the shortage to post-pandemic demand, aging infrastructure, workforce constraints, raw materials and product variation. Data centers add demand to a shortage that started earlier.
Open report card βAI-contributingβAI is making ordinary memory and storage more expensive.β
Documented supply competition; consumer surcharge unmeasuredMicron reports that HBM uses about three times the production capacity of DDR5 and that AI and conventional server demand are tightening DRAM and NAND supply. Available price series do not isolate a consumer surcharge.
Open report card βScapegoatedβEvery AI prompt uses a bottle of water.β
There is no reliable amount that applies to every promptMeasured water use varies by more than 10,000-fold across workloads. Server efficiency, utilization, cooling, climate and the source of electricity all change the result.
Open report card β