AI is everywhere. Since ChatGPT reshaped the conversation around AI but also it’s real-world applications in 2023, it has become ubiquitous. While skepticism about AI’s capabilities and impacts remains, experience continues to show that for many workers – including federal agency employees – generative and agentic AI can improve how work gets done and strengthen the delivery of mission-critical services and it’s ready for use now. The key issues now are ensuring that both agency workers and systems are ready to work with AI. In a new video conversation from GovCIO Media and Research, leaders from the National Institutes of Health, the Department of Education, and a trusted government partner, Maximus, discussed how they are addressing workforce readiness and system security to enable the rapid deployment and adoption of valuable AI tools and solutions.
As Mahua Mitra, Chief of Enterprise Platform Services, CIT at the National Institutes of Health, and Davon Tyler, Acting CTO and CISO, Office of Federal Student Aid at the Department of Education, both noted during the conversation, ensuring workforce AI readiness is one of the most important functions for federal agencies and their IT leaders today. With competition for AI-ready workers at an all-time high, the smart strategy for federal agencies is to train their existing workforce in the fundamentals of AI and create an environment where learning never stops.
It should come as no surprise that Taylor and Mitra both shared that workers can be nervous when AI is discussed. “There is the fear that AI will replace them in their jobs,” he shared. “People seem okay with generative AI and using AI for research, but agentic AI is a cause for concern.” For Featheringham, one of the other key components in getting a team AI-ready is to be transparent about potential problems or challenges that workers may face using AI and the tools they’ve been given. “Don’t hide the faults,” she advised, “be transparent and help team members to solve them and learn.”
There are, of course, many other requirements to take into consideration when equipping the federal workforce for AI. From ensuring data is cleansed and accurate so that there is no misinformation from chatbots or hallucinations from research to identifying a strategy, be that use case-based or enterprise-wide deployment, Mitra, Tyler, and their counterparts across the federal government have many factors to consider with AI deployment. However, at the top of the list, beyond even securing worker buy-in, is ensuring that the AI agency workers are using is secure.
The on-going focus on Zero Trust has improved federal agency network and application security, but as Featheringham explained, humans can often be the greatest source of security risks. To mitigate not only human security risks, but all forms of security risk, Featheringham recommends starting with a full understanding of the workflow for applications, generative AI tools, and agentic AIs to identify possible vulnerabilities on the technical side. Then, using cybersecurity best practices, plus the knowledge of how AI is being used to understand and address human-derived vulnerabilities. Only once risk is understood from both sides can “governance from the people and technical sides” be achieved, noted Featheringham.
With AI just beginning to disrupt how federal agencies work and deliver on their missions, conversations around AI will evolve as quickly as the technology. For agencies getting started on their AI journeys today or perhaps switching from a use-case-based adoption to an enterprise strategy, beginning with this conversation is the perfect starting point to build knowledge and ensure they’re ready to equip their federal workforce to be ready for AI.
Let’s get started! Access the conversation here.