OMB is releasing the consolidated AI Use Case Inventory for 37 federal agencies, collecting more than 1700 publicly reportable use cases in various stages of development and deployment to support agency missions.
A few things I want to highlight as we look to forward to seeing folks dive into this dataset:
1 – Probably one of the more extensive data calls regarding AI use in federal government yet. But you can’t improve on what you don’t measure! For agencies looking to mature as an enterprise around AI, we collected as much information as we could for use cases that were making the most of the infrastructure around them. Documentation? Evaluations? Point solutions that scale? We wanted to give agencies the opportunity to report not just on how they are using AI, but how they have matured their IT infrastructure to support governance requirements too.
2 – For the general public, and AI policy makers alike, it can still feel like a mystery as to how AI use cases and projects within agencies even get started. We asked agencies to provide information for things like contract IDs and system names for associated ATOs because the reality is many use cases rely on existing contract vehicles to get going. That has huge implications for innovation.
3 – A data call like this as a first go around is incredibly difficult to pull off. Without many of the guiding policies put in place such as M-24-10, M-24-18 and Executive Orders on AI put out by previous and current administrations – much of the information simply would not have been possible to collect. Every record in this dataset represents a significant time investment from agencies doing the work to meet with use case owners, stakeholders, privacy experts, procurement officials.
Not Yet a Premium Partner/Sponsor? Learn more about the OS AI Premium Corporate and Individual Plans here. Plans start at $250 annually.