Stanford's AI Index: adoption and capability surged, trust and readiness did not
The field's most comprehensive annual read found 78 percent of organizations using AI in 2024, up from 55 percent a year earlier, and 252 billion dollars of corporate investment. Capability jumped on every benchmark. Trust, regulation, and workforce readiness did not keep pace.
Stanford's AI Index, the most cited annual survey in the field, reports organizational AI use rose to 78 percent in 2024 from 55 percent the year before. Corporate AI investment hit 252 billion dollars, with US private investment at 109 billion. Model capability surged across hard benchmarks and the cost of running models fell steeply.
The gaps are in the human layer. Public trust is uneven and often falling in wealthy countries, regulation lags the technology, and adoption is running ahead of readiness.
The machines got better and cheaper fast, and companies rushed in. What did not scale at the same rate is the human side: trust, governance, and the capability to use the tools well. The bottleneck has moved from the model to the organization.
The constraint is no longer access to good AI. It is the readiness of the people and processes around it.
Adoption running ahead of readiness is a risk, not a win. Broad use without capability is where value leaks.
Cheaper, stronger models mean 'wait for the tech' is no longer an excuse. The work is organizational now.
Assume the model is good enough. Spend your effort on capability and governance, not on picking a model.
Pair every rollout with a readiness check: do people know how to use it, and can you see what they do with it.
Budget for trust. Clear rules and visible controls are what let adoption turn into value.
Benchmark gains are not business gains. Capability on a test is not capability in your workflow.
The adoption figure counts any use, however shallow. Do not read 78 percent as 78 percent transformed.