HHS’ 2024 AI use case inventory shows move toward internal chatbots

The Department of Health and Human Services reported the most artificial intelligence use cases of any federal agency in 2024, with the agency’s inventory of reviewed AI solutions showing a growing embrace of chatbots for internal and external operations. 

In a Dec. 16 blog post, Steven Posnack — HHS principal deputy assistant secretary for technology policy and principal deputy national coordinator for health IT — said that “in comparison to our 2023 AI use case inventory, which included 163 use cases, the 2024 inventory now includes 271, representing a 66% increase.” 

Federal guidance has sought to promote greater transparency around agencies’ evaluation and adoption of AI technologies in recent years, with their use cases and related risks increasingly requiring public disclosure. 

Then-President Donald Trump first issued an executive order in December 2020 that required agencies to compile and report inventories of their AI use cases. A subsequent executive order released by President Joe Biden in October 2023 expanded those disclosure requirements, and guidance released by the Office of Management and Budget in March 2024 further mandated that agencies note which of their use cases are safety- and rights-impacting. 

As of Dec. 16 — the mandated deadline for required federal entities to submit their inventories — 37 agencies reported more than 1,700 AI use cases, according to OMB’s Github repository. HHS reported the most AI use cases, according to the Github compilation, followed by the Department of Veterans Affairs with 229 — though VA’s public-facing inventory lists 227 use cases. 

Posnack said in his blog post that, of HHS’ 271 use cases, “59 are ‘Initiated,’ 57 are in ‘Acquisition and/or Development,’ 35 are in ‘Implementation and Assessment,’ 104 are in ‘Operation and Maintenance,’ and 16 are ‘Retired.’”

Of particular note in the agency’s 2024 inventory is the number of use cases related to chatbots, with more than two dozen of the reported solutions looking at using generative AI to streamline public engagement with the department and to help its employees better access and retrieve information. 

By comparison, its 2023 AI use case inventory reported seven total instances where the department was exploring the use of chatbots. 

The most recent AI use cases that HHS reported include an initiated effort to create a chatbot that can help with applying for grants, an initiated chatbot to help Division of Global Migration Health personnel “with developing an initial draft response to inquiries where the response could have been found on our website” and a chatbot to help researchers find data sets for environmental health research efforts, which is in the acquisition and/or development phase.

Seven of the listed use cases pertain to the development of a GenAI chatbot for the Centers for Disease Control and Prevention, which is called ChatCDC.

The tool, which HHS said is powered by “[Microsoft] Azure OpenAI Large Language Models,” is for internal agency use and is designed to provide responses for a variety of tasks, including extracting information from larger pieces of text, generating software code and recommending edits to documents. 

Five of the ChatCDC’s use cases are in the operation and maintenance phase, while the other two are in the implementation and assessment phase.

“The chatbot's output can then be copied and pasted or manually entered into the appropriate testing process by the staff to ensure any code used from the chatbot's output meets the criteria of the current task,” HHS said in its AI inventory.