The two kinds of AI for government: Everyday and transformational
You want to know more about how AI can work for your agency, but you aren’t sure if you’re ready for a technological overhaul to implement it. Thankfully, there are different levels of AI usage—here's how to tell which is right for you.
Agencies and organizations are increasingly exploring and adopting artificial intelligence. As the options and capabilities grow, the possibilities can feel overwhelming—especially when you consider the security requirements, the breadth and depth of agency data, and the necessary integrations and training.
But there’s an easier way to think about AI. We’ve already been using it for a while: New AI innovations, including the latest generative AI (Gen AI), are simply building on the older technology that we’re already used to. It can be helpful to break AI into two categories: The everyday, and the transformational. Here’s what they are and how to know which is right for your use case.
Everyday AI vs. transformational AI in government
Everyday AI operates on a small scale, focused on productivity. It’s the established AI we’re already using to assist tasks and help us shift our own focus from repetitive tasks to more creative ones. Autocorrect, automated workflows, chatbots, and increased personalization are all examples of everyday AI.
Transformational AI operates on a much bigger scale. Instead of supporting individual tasks, it has the potential to change an entire enterprise or industry. Transformational AI can evolve organizational capabilities, create new results on its own, and enable us to discover insights we wouldn’t have otherwise. For instance, in regulatory compliance, transformative AI is being used to expedite review of regulatory comments. This application of AI parses through vast quantities of public feedback, identifying key themes and concerns with efficiency that surpasses manual methods.
Getting started
Everyday AI is the best way to start, if you’re not using it already. It will give you and your workforce the opportunity to get comfortable with AI in an easier, less seismic way, learning its possibilities and limitations with fast results and individual impact.
Whether you’re already using everyday AI or not, focus on upskilling your workforce and training them to use it across tasks. Your agency can simultaneously assess and prepare its data for transformational AI in the long-term.
Before embarking on this journey, pinpoint the problems you aim to solve with AI. Set aside the allure of a broad tech sweep and concentrate on specific areas—be it enhancing the citizen and employee experience, refining business functions like accounts payable, or optimizing processes from procurement to retirement.
Good starting use cases for everyday AI improve workforce productivity. Automating repetitive tasks, particularly ones that use templates, are perfect for square one. AI that’s secure at the enterprise level can pull content into a standardized format, extract basic data from spreadsheets, read and interpret files, and change one file type to another according to your specifications. It can also be used to improve more creative projects: For example, we’re using AI to help our clients’ HR departments attract and retain high-quality talent in today’s competitive labor market by creating skills assessments with greater speed, efficiency, and accuracy.
"If you’re looking ahead to transformational AI, data analytics will be an early area of opportunity."
Data needs to be prepared—organized, standardized, duplicates removed, etc.—before it can be read and interpreted by AI. But it may soon be possible to use one AI to prepare your data so another AI can analyze it for you. Once the data is analyzed, we may not submit it to a dashboard like we do now, forcing users to adapt their questions and behavior to the design and data interpretations available. Instead, we could replace dashboards with advanced Gen AI chatbots able to adapt their answers to the user’s question without sacrificing accuracy or security.
Because everyday and transformational AI build on what’s already available, there are experienced partners who know how to narrow a use case, identify how rigorous of a solution is needed, customize it, implement it, and provide support throughout any necessary iterations, training, and change management. At ICF, we are using AI to optimize productivity and spur innovation while helping our clients advance mission outcomes—and we find that this distinction between everyday AI and transformational AI is a helpful way to frame and scope the exciting possibilities.
