Halloween brings plenty of monsters into neighborhoods across the Greater St. Louis region. Most are kids in costumes looking for candy. The real threats to local government are usually much less obvious.

Cybercriminals are using AI to create polished emails, realistic voice messages, convincing videos, and believable requests. These attacks may look like they came from a department head, elected official, vendor, resident, partner agency, or utility customer.

For municipalities, public works departments, libraries, parks departments, utility districts, and economic development organizations, the stakes are different than a typical business. You are protecting citizen data, taxpayer resources, public trust, and the continuity of services people rely on every day.

The good news is that your staff does not need to become expert AI detectives. Your organization needs clear security rules that do not depend on someone noticing a strange voice, a small wording mistake, or an email that just feels off.

AI shapeshifters: Sounding official is not enough

Voice calls and video messages used to provide some confidence that you were dealing with a real person. That is no longer enough. AI can help scammers copy a voice, create a realistic video, or write a message that sounds exactly like someone your team knows.

A finance employee in Clayton, a public works coordinator in St. Charles, a parks manager in O’Fallon, or a library administrator in Edwardsville may be careful and still be fooled. AI-generated content is improving quickly, and many common warning signs are already outdated.

That is why sensitive requests need a verification process. Require a second form of confirmation before releasing funds, changing vendor banking details, sharing protected resident information, approving payroll changes, or granting access to municipal systems. Use a known phone number or another trusted communication method. Do not rely on the contact information included in the message.

Multi-factor authentication also adds an important layer of protection. A convincing message may fool someone, but it cannot bypass a strong process as easily. This is especially important for Microsoft 365 accounts, email, finance systems, permitting platforms, utility billing systems, and remote access.

AI mummies: Old scams with better disguises

Phishing emails used to be easier to spot. Poor spelling, awkward grammar, generic greetings, and strange formatting often gave them away.

AI has changed that. A phishing message can now be clear, professional, and personalized. It may reference a real city project, an open records request, a grant deadline, a vendor invoice, a council meeting, or a department name. Good writing does not make a request safe.

Train your team to evaluate the request instead of only evaluating the message. Is someone asking to change payment instructions? Is an employee requesting sensitive information from a personal email account? Is a vendor pressuring your staff to skip the normal approval process? Is someone asking for resident data, utility account information, employee records, or access to a shared mailbox?

Those questions matter more than perfect punctuation. A scam can be well written and still be a scam.

This applies whether you serve residents in Chesterfield, Maryland Heights, Belleville, Collinsville, or a smaller community in the Metro East. Attackers do not care about city boundaries. They care about finding a process they can exploit.

AI vampires: Do not invite them into public data

Unapproved AI tools create another risk. Staff may use free or public AI services to summarize reports, draft resident communications, analyze spreadsheets, organize meeting notes, or help with grant language. These tools can be useful, but they can also expose confidential or regulated information.

Before using an AI tool with government information, your team should know where the data goes, how it is stored, who can access it, and whether the provider uses it to improve its service. If no one can answer those questions, the tool should not receive protected information.

Create a simple approved tools list and give staff clear examples of what they can and cannot enter into an AI system. Your policy should address citizen information, utility account data, police or court-related information, personnel records, vendor contracts, building plans, passwords, financial data, and anything tied to public safety or critical infrastructure.

For many public agencies already using Microsoft 365, this is also a good time to review security settings, data retention, conditional access, sharing permissions, and how AI-related tools are governed. The goal is not to block useful technology. The goal is to use it in a way that protects public trust.

The best defense against AI threats is a better process

AI has made scams cheaper to create, faster to launch, and harder to recognize. Your best defense is not asking every employee, department head, clerk, technician, or board member to become a cybersecurity expert. It is building simple habits that protect your organization even when a message looks completely legitimate.

Start with multi-factor authentication. Use password managers. Keep systems and devices updated. Back up important data. Train staff to question unusual requests. Require verification for financial and account changes. Review who has access to sensitive information. Document your process so it survives staff turnover, elections, retirements, budget cycles, and emergency situations.

Strategic technology planning matters here. Local governments across St. Louis are expected to deliver more digital services, respond faster, protect more data, and do it all with limited staff and taxpayer-funded budgets. A practical security plan helps you prioritize what matters most without chasing every new headline.

These steps may not feel as dramatic as fighting a monster, but they can prevent a costly incident, a service disruption, a public records problem, or a loss of resident confidence. At Tigerhawk, we help local governments and public agencies put practical security rules and technology in place without making daily work harder.

Need help protecting your organization from AI powered scams? Schedule time with Tigerhawk to talk through your next steps.

Common questions from St. Louis public sector leaders

How can a St. Louis municipality protect citizen data from AI phishing attacks?

Start with the basics that reduce risk quickly. Require multi-factor authentication, verify financial or data requests through a trusted channel, limit access to sensitive systems, and train staff on real examples they may see in city operations. Then review Microsoft 365 permissions, shared mailboxes, and vendor access so citizen data is not exposed unnecessarily.

What should public works or utility departments in Greater St. Louis do about AI voice scams?

Create a rule that voice alone is never enough for sensitive action. If someone requests a payment change, emergency purchase, account update, or system access, staff should verify it using a known number or established internal process. This is especially important for utility billing, field operations, infrastructure vendors, and after-hours service situations.

Do libraries, parks departments, and economic development agencies need a government AI policy?

Yes, even if the policy is simple. Staff may use AI to draft content, summarize reports, or organize program information, but they need clear limits around resident data, personnel records, grant details, contracts, and confidential planning documents. A practical AI policy helps departments use new tools responsibly while protecting public trust and taxpayer resources.