12 Graphs That Explain the State of AI in 2026

The buried story in this year's Index isn't capability benchmarks — it's the municipal datacenter restrictions, a regulatory pressure point that rarely surfaces in coverage dominated by model releases and funding rounds. That's a land-use and infrastructure constraint that could matter more to build-out timelines than any federal AI policy currently in motion.

This report lands on the same day MIT Technology Review published its own chart-driven breakdown of the Index ("Want to understand the current state of AI?"), which means the data is getting wide distribution but the interpretive frames vary significantly. Where MIT TR focused on jobs and capabilities, the IEEE Spectrum framing emphasizes financial and political friction — IPO pressure from OpenAI and Anthropic, and public backlash. That backlash theme connects directly to "Why opinion on AI is so divided," also from MIT TR on April 13, which used the same Index to examine fragmented public trust. Taken together, the three pieces suggest the Index is functioning less as a neutral scorecard and more as a Rorschach test for whoever is reading it.

Watch whether OpenAI or Anthropic file S-1 disclosures within six months of this report's release — if either does, the Index's IPO framing will look prescient; if neither moves, it signals the public backlash data is weighing on investor appetite more than the capability gains can offset.