As artificial intelligence becomes embedded in public services, communications, and decision-making, governments across the United States are moving quickly (but not uniformly) to regulate its use.
For municipalities and public agencies, this creates a complex and evolving compliance landscape that touches everything from transparency requirements to liability for misinformation.
The fear of liability, along with not truly understanding the capabilities and reality of AI, is what keeps many municipalities from taking advantage of this technology.
A Patchwork of AI Laws
AI regulations in the United States are not governed by a single federal framework. Instead, states are defining their own rules, often focusing on specific use cases like hiring, deepfakes, consumer protection, or automated decision-making.
This creates more than just variation. It creates inconsistency in compliance expectations.
What’s permissible in one state may require disclosure (or be restricted entirely) in another. And even within a single state, requirements may differ depending on how AI is used.
States without formal AI-specific laws are still enforcing AI-related issues through existing consumer protection, privacy, and anti-discrimination statutes.
For local governments, this means compliance is no longer just about what tools you use—it’s about how those tools are applied across different contexts.
AI Transparency and Disclosures
Across nearly all emerging AI laws, one principle is consistent: users should know when AI is involved.
Disclosure requirements are increasingly applied to:
- Chatbots and virtual assistants
- AI-generated summaries or translations
- Automated decision-support tools
In practice, this means governments must clearly communicate when a resident is interacting with AI; when content has been generated or assisted by AI; and when AI plays a role in decisions that affect the public.
This is not about over-explaining but maintaining trust through clarity.
Risks and AI Hallucinates
One of the most pressing challenges with AI systems is their ability to produce hallucinations—outputs that are incorrect, fabricated, or misleading, yet presented as fact.
For governments, this is not a theoretical issue. It has direct implications:
- A chatbot provides incorrect permit information
- A document summary misrepresents a policy
- A generated response cites non-existent sources
These failures create risk across three dimensions.
1. Legal Risk
Courts have already begun enforcing this—particularly in legal settings where AI-generated inaccuracies have led to sanctions.
- Consumer protection (misleading information)
- Professional liability (failure to verify)
- Fraud or misrepresentation (false claims presented as fact)
2. Operational Risk
AI errors can disrupt services, create confusion, and increase staff burden as teams respond to misinformation.
3. Public Trust Risk
When government information is wrong, even once, it erodes confidence quickly—and is difficult to rebuild.
Protecting Your Municipality
AI hallucinations are not governed by new laws—they are being enforced through existing ones. Whether framed as consumer deception, professional negligence, or misrepresentation, courts are already holding individuals and organizations accountable for AI-generated falsehoods.
Here are ways to protect your municipality:
Provide Human Verification of AI Outputs
This one is simple: Do not publish AI-generated content without human review and approval. This includes website content; public notices; and policy summaries.
- Add a “Verified by [Role/Department]” checkpoint before publishing
- Use AI for drafting only—not final output
- Create a verification checklist, such as, “Are facts sourced and accurate?” “Are dates, names, and links validated?” and “Does this align with official policy language?
Establish Accountability Standards
This is where most organizations fall short. Accountability requires clear ownership, not shared ambiguity.
Content Ownership (Internal)
Assign a department owner for every AI-assisted output. For example:
- Clerk’s Office manages meeting summaries
- Communications manages website content
System Oversight (Administrative)
Designate an AI governance lead (or equivalent role) who will:
- Approve AI tools
- Define acceptable use cases
- Maintain documentation
Vendor / Platform Responsibility (External)
Require vendors to:
- Document how AI systems function
- Disclose limitations and risks
- Maintain security and compliance standards
Fragmented systems, disconnected tools, and plugin-based environments make it difficult to maintain consistency, ensure accuracy, and assign accountability
When AI is layered onto an already complex system, risk compounds.
A More Structured Approach to AI in Government
This is where system design becomes critical.
The Local Level platform is built as a connected system—not a collection of tools, plugins, or patches. Our structure directly supports responsible AI use by:
- Embedding approval workflows into publishing
- Aligning content with department-level ownership
- Centralizing updates so corrections apply everywhere
- Structuring information to reduce misinformation risk
Rather than relying on add-ons or plugins, AI operates within a framework designed for clarity, compliance, and control.
AI does not remove liability.
It redistributes it.
AI in government is no longer optional. But neither is accountability.
As regulation continues to evolve, three expectations are already clear:
- Transparency: Residents should know when AI is being used
- Accountability: Governments remain responsible for outcomes, even if AI is involved
- Risk Management: High-impact systems require oversight, testing, and documentation
The challenge is not whether to adopt AI.
It’s whether your systems are built to support it responsibly, reliably, and at scale.
Click here for an examples of regulations by State
California: Broad transparency, deepfake, and AI governance laws
View laws ➡️
Colorado: Comprehensive “Colorado AI Act” (risk + discrimination focus)
View laws ➡️
Texas: Responsible AI Governance Act (enterprise oversight)
View laws ➡️
Utah: AI Policy Act (consumer-facing AI disclosures)
View laws ➡️
Illinois: AI use in hiring + biometric/algorithmic regulation
View laws ➡️
New York: Bias audits + transparency (especially in hiring tools)
View laws ➡️
Virginia: AI-related provisions tied to privacy and platform regulation
View laws ➡️
Florida: Emerging “AI Bill of Rights” and content regulation efforts
View laws ➡️
Georgia: AI chatbot safety and liability legislation
View laws ➡️
Alabama: AI safety commissions + chatbot safeguards
View laws ➡️
Oregon: AI enforcement through consumer protection actions
View laws ➡️
New Jersey: Civil rights + algorithmic discrimination oversight
View laws ➡️
Massachusetts: AI enforcement via privacy and misrepresentation laws
View laws ➡️
Washington: Early adopter of deepfake and biometric AI restrictions
View laws ➡️