Inflation.

Four of eight questions produced clear consensus, including three damaging 5-0 verdicts where all models confess that the US government does not accurately measure inflation and that neither CPI, PPI, nor PCE stands as the definitive best measure. Grok stands alone in denying that Trump's tariffs raised grocery prices, a position rejected by every other model. The sharpest fault lines emerge on Federal Reserve independence, where GPT and Grok defend its autonomy, and on 2026 rate cuts, where Claude and DeepSeek break from the majority expecting reductions.