Today’s experts agree on the potential of Large Language Models (LLMs) in government agencies. Some benefits include improved knowledge management and a reduced burden on tedious tasks.

Andres Perez from the CMMC hits the nail when he states that an organization may have data that can answer questions they did not know existed. The CMMC has put together a large Knowledge Management Platform that assists federal professionals in accessing information.

Many federal agencies have sensitive data that they would like to generate insights from, but not have it exposed in an insecure manner. One method to do this ethically is to create “synthetic” data sets where conclusions can be drawn while never compromising personally identifiable information.

Rather than thinking LLMs are the end answer, many subject matter experts comment that AI and LLMs should be viewed as extensions of human abilities. Chris Roberts from Quest states that this technology should be viewed as an augmentation to strategy, not a replacement.

Developments in using AI to derive information from LLMs are changing rapidly. In addition to policy on data security, Don Wildner from BAE suggests that federal agencies may have to organize new policies on how to ethically create a prompt. Some may be allowed, some not.