Spotlight on U.S. Government Use of AI
Researchers from Stanford and New York University released a report this month with their findings of how the U.S. government is using artificial intelligence (AI). Here are their five high-level findings:
- The government’s AI toolkit is diverse and spans the federal administrative state. There is a lot of experimentation going on at various agencies. AI is used for an array of things, from enforcing regulatory mandates to assessing weather patterns.
- Despite wide agency embrace of AI, the government still has a long way to go. In most cases, the researchers found that the AI applications in use were not highly sophisticated.
- AI poses deep accountability challenges. Government decisions often require justifications and AI tools are generally not given statutory validity.
- If we expect agencies to make responsible and smart use of AI, technical capacity must come from within. Almost half of AI applications were developed by outside or contracted sources. The researchers assert that in-house development is preferred, but note that acquiring and maintaining expertise will be an ongoing challenge.
- AI has the potential to raise distributive concerns and fuel political anxieties. Nefarious actors could figure out the AI algorithms used and work to game the system, undermining public confidence.
For information on AI in the private security sector, see these Security Management articles: