USDA’s Measured Approach to Responsible Advanced Analytics
The Food and Nutrition Service is focusing on data before it scales AI and ML capabilities.
Through its 15 nutrition assistance programs, the Food and Nutrition Service (FNS) manages America’s nutrition safety net and serves one in four Americans a year — and it collects a lot of data. While FNS has built up a robust data infrastructure, it’s taking a measured approach to deploy responsible AI instead of jumping ahead to apply advanced analytics across all of its data sets.
The FNS’s Enterprise Data Analytics Platform (EDAP) suite includes a data lake portal and dashboarding capabilities. So far, these capabilities are being deployed cautiously across the service.
“We really laid the groundwork — at least in terms of the technology — we have the capability to utilize artificial intelligence and machine learning,” said Chris Rottler, FNS Assistant Chief Data Officer, at ATARC’s AI summit on Tuesday. “The question we’re really tackling now is when and where does it make sense to employ these methods when we have that data available, and how do we do it in a responsible manner?”
For Rottler, this means focusing on the data.
“It’s all about the data and understanding the context and digging into the structures that generate the data,” Rottler said. “Inequality is insidious and not always easy to uncover in data and the questions that we ask. The onus is on us to ensure that the fuel that powers the models we develop first, does no harm, and second, does not exacerbate inequality.”
By teaming up data and policy experts, the service is taking a collaborative approach to understanding and improving its survey practices.
“Pairing our data experts with our policy experts in [the Supplemental Nutrition Assistance Program (SNAP)] has ensured that we’re really asking the right questions and that we’re able to draw accurate conclusions,” FNS Senior Technical Advisor Christine Daffan said.
FNS is looking at data analytics to support not only its external programs but also its internal operations; including staff diversity, equity and inclusion.
“We’ve created a barrier analysis dashboard that looks at the landscape of our staff — who do we have on board doing this work, providing these services — and where there are opportunities for the agency to increase our diversity, and also the accessibility,” FNS Assistance Administrator Tameka Owens said.
Rottler said that, although the FNS is just dipping its toes into advanced analytics, he is enthusiastic about all of the groundwork they have done to prepare for next steps in responsible AI.
“I know I sound really cautious, but I am really excited for FNS and USDA as a whole,” Rottler said. “We are positioned to begin to utilize machine learning more in a responsible way. And we have piloted its use in a very controlled and specific use case. … We’re taking baby steps, but I’m excited to see how we can really take advantage of the investments we made in the infrastructure.”
This is a carousel with manually rotating slides. Use Next and Previous buttons to navigate or jump to a slide with the slide dots
-
Cyber Resilience and Recovery Amid Evolving Cyber Threats
Data durability is a key aspect of NIST’s cybersecurity framework for public and private organizations.
21m listen -
The CAIOs Leading Responsible AI Development Across Government
Since the White House's AI executive order, federal agencies are in the process of naming chief artificial intelligence officers.
7m read -
Defense Board to Pitch Solutions for Closing Tech Talent Gaps
Defense Innovation Board members cite need to modernize people management the same way government modernizes technology.
4m read -
Energy Researchers Aim For Holistic Approach to AI Issues
A new center at the Oak Ridge National Laboratory is looking at under-researched areas of AI to better understand how to secure it.
2m read