Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Technology Policy topic
No spam. Unsubscribe anytime.
County staff warns against feeding privileged data into public AI platforms; proposes unified AI policy and training
Summary
A county technology lead told commissioners to avoid inputting privileged personal data into open AI platforms, recommended county-specific models and uniform policies, and signaled additional training and a potential closed-session review for cybersecurity topics.
S6, a county technology lead, told the court AI is "a way of the future" but warned that staff must not input privileged data—driver's licenses, Social Security numbers or other sensitive records—into open, third-party large-language models where the data could be retained and exposed. He said the county is completing DIR-required AI training and proposed a single county policy covering AI and cybersecurity needs across departments.
S6 said the county could develop smaller, county-only language models that never transmit local data to public platforms and urged that department heads help enforce policy. He recommended discussing security-sensitive AI topics in closed session to avoid broadcasting defensive posture and vulnerabilities. "Not to put any privileged data into AI services," S6 said, summarizing his central caution to staff.
