Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Technology Policy Ai topic
No spam. Unsubscribe anytime.
County adopts initial artificial intelligence policy after IT flags malicious AI bot
Summary
Bannock County reviewed a draft AI policy, discussed recent detection of an AI-based Trojan, and authorized adoption of a first-draft policy to govern county use of AI tools.
Adam McKinney, chief technical officer, presented a first draft of a Bannock County artificial intelligence policy and recommended adopting a cautious approach to county use of AI.
McKinney said the draft policy was developed using sample policies from the state of Washington, Sonoma County (Calif.), and NACO guidance, and he told commissioners the county’s new SentinelOne antivirus had detected an AI-based Trojan called "Just Ask Jackie" installed on a prosecutor’s laptop. McKinney described the Trojan as software that "installs on the machine and then it expedites, malicious code at random times," and said that detection underscored the need for policy and controls.
The proposed policy would permit vetted AI tools in controlled ways, require human review of AI-generated materials before release, and include specific language protecting county trademarks and seals from unvetted AI-generated content. McKinney said the policy lists approved vendors (for example, Google, Microsoft, Apple, Wolfram Alpha, LexisNexis) and excludes tools his team judged biased or unsuitable. He also described options such as web filtering and sandboxed, auditable AI tools, and said vendors had not yet provided a reliable audit trail to prove a dataset has not changed.
Commissioners discussed enforcement, training and placement of the policy in personnel rules; one commissioner moved to adopt the policy and the board approved the recommendation by voice vote. McKinney and county HR will coordinate rollout, include the policy in the personnel policy between the social media and IT sections, and brief department heads.
