Five AI use cases, ready-made for public agencies

Known as “the crossroads of the nation” for its proximity to some of the country’s busiest transportation routes, the small Missouri city of Wentzville could just as easily have earned that nickname for staking out new territory at the intersection of artificial intelligence (AI) technology and the citizen experience.

Wentzville, a town of about 47,000, is using generative AI (genAI) tools to streamline communications with its citizens, according to reports, putting it in a small minority of municipalities that are actively using genAI.

AI gives cities a powerful new set of tools to fulfill what has been a longstanding goal for many: to enrich the citizen experience. According to McKinsey & Co., citizens are nine times more likely to trust a government agency when they are satisfied with the service they receive. It’s incumbent on public agencies to explore how AI can benefit their constituents, posits Anthony A. Fisher, head of data governance and artificial intelligence with Colorado’s Department of Revenue. “I feel strongly that AI holds the key to unlocking a plethora of services that otherwise may be impossible to implement within government.” The challenge is identifying use cases that are likeliest to deliver value, without exposing the agency or its constituents to undue risk.

Based on my experience supporting public agencies in their AI journeys, here are five of the most compelling government use cases for the technology today:

1. Creating a single, 360-degree profile of citizens so agencies can better understand individual citizens’ preferences and needs. As part of an effort to modernize how it interacts with citizens, the city of Christchurch, New Zealand centralized customer data to create a single, unified identity and view of each of its citizens. To do so, it began shifting away from the siloed legacy systems on which it had been relying in favor of a centralized, AI-supported system for managing data. Now citizens will enjoy the convenience of using a single digital identity to access city offices and services whether interacting as a citizen or a business.

2. Predicting citizen needs to suggest and provide personalized services. With a 360-degree view of its citizens, Christchurch can offer highly personalized service recommendations generated by AI models, like how AI generates tailored suggestions for retail customers. Using AI tools, agencies can automate delivery of personalized recommendations and communications, such as providing parents with information about school registration, parks & recreation programs, and summer camps for their kids. During interactions with citizens, genAI can feed service representatives relevant in-the-moment information about the person with whom they’re interacting, along with suggested best next actions, to help resolve an issue. Intelligent assistants also can translate agency communications into multiple languages and know exactly whom among their citizenry to send those communications, in which language. Sacramento, Calif., is using an AI-powered translation system to provide translated text and speech to citizens in real-time during city agency meetings.

3. Simplifying, accelerating and automating services with a citizen-friendly, mobile-enabled platform. Giving citizens access to a mobile-enabled, genAI-supported digital platform to connect with an agency can make once-dreaded interactions with agencies a breeze by gathering information from citizens (with the help of document information extraction using optical character recognition), recording it in a form and database, then processing it to trigger an outcome, such as renewing a vehicle registration or a driver’s license.

Natural language processing can streamline application and service request journeys, enable frictionless self-service for basic inquiries, and point users toward the right agency resources, regardless of the communications channel they’re using. This in turn relieves service teams of the tedious, time-consuming processes involved in processing and developing customized, manually created responses to inquiries. Instead, AI automatically categorizes, routes and responds to various types of inquiries and issues, with templates to ensure consistency of responses to simple inquiries, while also identifying and routing cases that require human intervention for resolution.

Mobile, multichannel access is critical to the usability of platforms like this. In San Diego, residents can use the city’s “Get It Done” app to report non-emergency issues like potholes or graffiti, with the ability to upload a photo in real time to document the issue. As one reviewer described it, the app is “much better than filling a form out on some obscure city website, drafting an email or god forbid, having to call someone.”

4. Handling mass requests, during a crisis and otherwise. When COVID-19 hit, the city of Hamburg, Germany, volunteered its services to the German finance ministry, offering to quickly develop and launch a platform for processing aid applications from artists across the entire country who were struggling financially because of the pandemic. Three weeks later, the platform, driven by AI-based decision support and automation, was live, rapidly, securely and accurately evaluating and classifying some 2.7 million documents from a wide range of sources and trafficking applications from initial filing to payout. All told, those AI capabilities saved an estimated 13,000 to 33,000 hours of manual work. The platform performed so well, in fact, that it’s being adapted to other use cases in Hamburg and across Germany.

5. Listening and taking action to consistently improve the citizen experience. One of AI’s strengths is conducting sentiment analysis of opinion data from structured and unstructured sources (surveys, social media, etc.), then sharing insights that agencies can act upon to improve the citizen experience.

As promising as AI use cases like this are, their success depends largely on solid preparation. That includes ensuring your AI software meets strict cybersecurity standards and is being fed relevant, high-quality data, and taking a measured approach by piloting AI in targeted use cases before integrating it more broadly. You’ll also need to enact a clear set of AI ethics and governance policies and provide training to employees on how to use the AI tools at hand. And finally, seek out support from your peers. Groups like the GovAI Coalition, formed last year by the city of Sacramento, are giving representatives from local, county and state governments across the country a forum to help one another succeed with AI.

Based in Sydney, Australia, Kathleen O’Brien is a product marketing expert, and the global public sector customer experience lead at SAP. She’s a cross-industry expert with 25+ years global experience across public sector, retail B2C/B2B, and consumer products. She has a deep desire to improve the way people interact with government to achieve better outcomes for all.