Cautious approach

USDA Sees Plenty of Generative AI Uses But Is Careful on Its Tool Picks

Generative artificial intelligence offers plenty of potentials for the U.S. Department of Agriculture, but a cautious approach is needed because of the uncertainties in its tools, a USDA official said.

Christopher Alvares, USDA chief data officer, told NextGov/FCW that while generative AI can bring more efficient and more effective USDA programs, the department is weighing questions on its efficacy and possible threats to security and privacy.

He pointed out some of the challenges in AI use, such as addressing its tools’ data quality issues, particularly unwanted biases.

Jim Barham, assistant chief data officer at the USDA Rural Development Innovation Center, shared that much data foundational work is now being done on rural development, one of the department’s eight mission areas.

According to Alvares, the USDA currently uses AI in conservation practices analyses, determining forest health and wildfire risks through satellite imagery and predicting invasive species’ spread.

The department’s 2023 public inventory of AI use cases listed 30 items, NextGov/FCW reported.