Secure AI for Local Government vs. Public AI

Artificial Intelligence (AI) is rapidly becoming a part of everyday work. Public AI tools like ChatGPT, Claude, and CoPilot are helping people summarize information, draft documents, analyze data, and answer questions faster than ever before.

As AI becomes more popular, local governments are asking an important question: “Is it safe to use AI with sensitive municipal information?”

Local governments routinely handle sensitive information, including Social Security numbers, driver's license numbers, utility account data, employee records, and law enforcement information. Before introducing AI into daily operations, municipalities must understand how these tools handle and protect their data.

Understanding the following three areas can help local governments make informed decisions about AI:

1. How the AI was trained
2. Where the AI gets its answers
3. What happens to municipal data after it is uploaded

To evaluate an AI solution's security and reliability, local governments must first understand how it was trained. Public AI solutions are trained on a combination of information from the internet, third-party sources, and user data.

Municipalities should also understand where an AI solution is developed and hosted. Agency data is most secure when the solution is developed and hosted domestically. Even when a company is headquartered in the United States, it may outsource development overseas. This complicates data access and handling and opens the door for municipal data to be exposed internationally.

iWorQ’s AI tool, called XworQ, was intentionally developed in the United States (U.S.) with the needs of local government in mind. Our team in Northern Utah has trained and maintained it with a focus on the security, privacy, and operational needs of local governments.

Once an AI has been trained, the next question becomes: where does it get its answers from? Public AI tools draw on a vast knowledge base built from information across the internet. This broad foundation makes them powerful tools for answering questions on a wide range of topics.

Public AI falls short in its lack of agency-specific knowledge. Because these tools rely primarily on general information, they often provide general answers. A local government employee may be able to improve the quality of those answers by uploading agency-specific documents, but doing so places municipal information into a public AI platform that collects user data.

XworQ takes a different approach. It safely allows agencies to upload documents, ordinances, policies, and other municipality-specific resources, providing accurate, relevant answers based on the agency's own information. For example, a code enforcement official can quickly locate the section of an ordinance related to noise complaints, while a resident can find the inspections required for a new building permit in their community.

The most important question local governments should ask is what happens to their data after interacting with an AI tool. Before adopting any AI solution, municipalities should carefully review the vendor's privacy policy and data handling practices.

The privacy policies of public AI tools like ChatGPT state that your data may be shared with vendors, marketing partners, service providers, and other third parties. Interactions and user data are also used to train future models for many AI solutions. Purchasing plans with enterprise security makes a significant difference in keeping your data private, but it is often expensive.

When local government clients interact with XworQ, their chats, documents, and user data remain within iWorQ’s secure environment. Their data is not used to train XworQ or future iterations. It is never shared with, or made accessible to, other clients. iWorQ provides this level of security and privacy in a solution designed specifically for the operational and budgetary realities of local government.

Understanding how an AI solution is trained, where it gets its answers, and how it handles data can help municipalities make more informed purchasing decisions. Before selecting an AI platform, local governments should consider asking vendors the following questions:

XworQ was thoughtfully built in the U.S. to help local governments take advantage of AI without sacrificing control of their information. Unlike public AI tools designed to serve a broad audience, XworQ was built specifically for municipal workflows and uses the agency's own documents, ordinances, policies, permits, and regulations to provide relevant answers.

Municipal data remains isolated, private, and inaccessible to other organizations, giving staff the confidence to use AI for everyday tasks while maintaining the security standards their communities expect.

XworQ currently provides clients with 4 AI assistants: