HB 178 limits how state and local government entities in Montana may use artificial intelligence systems. The bill defines “artificial intelligence system” broadly and then prohibits government entities and state officers from using AI for cognitive behavioral manipulation, discriminatory classification, malicious purposes, or surveillance of public spaces, with narrow exceptions for locating missing, endangered, or wanted persons and for uses allowed under existing continuous facial surveillance law.
The bill also creates disclosure and human-review requirements. If a government entity or state officer publishes AI-generated material that has not been reviewed by a human in an appropriate responsible position, the material must be labeled as AI-produced. Likewise, public-facing interfaces that use AI must disclose that use. In addition, any AI-generated recommendation or decision affecting a person’s rights, duties, or privileges and immunities must be reviewed and may be rejected or modified by a trained human official.
Impact
HB 178 adds new restrictions and procedural safeguards to Title 2 of Montana law governing state government operations. It affects state officers, public employees, and local or state government entities by limiting certain AI uses, requiring disclosure of AI involvement, and mandating human oversight for consequential decisions. The bill also creates an exemption for peace officers when disclosure would jeopardize investigations or public safety, and it cross-references existing law on continuous facial surveillance.
Sentiment
The bill appears to have broad overall support, passing the House and Senate with strong margins and several unanimous or near-unanimous votes in later stages. The voting pattern suggests general agreement on the need for guardrails around government AI use, transparency, and human accountability. The final concurrence votes were especially strong, indicating that most lawmakers accepted the amended version.
Contention
The main points of contention likely centered on how far to restrict government AI use and how much discretion to preserve for law enforcement and public agencies. The bill’s limits on surveillance, behavioral manipulation, and AI-based classification could raise concerns about operational flexibility, while the disclosure requirements may be viewed as burdensome in sensitive investigations. The peace officer exemption shows an effort to balance transparency with investigative needs, suggesting that law enforcement exceptions and the scope of prohibited AI uses were the most likely areas of debate.