Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Audit Technology topic
No spam. Unsubscribe anytime.
Internal audit plans broader data analytics and careful, sanitized AI use to expand coverage
Summary
Internal audit described its history with analytics tools (ACL to Arbutus), current use of Power BI and SharePoint dashboards, and controlled use of public-tier AI for scripting and drafting; staff emphasized manual sanitization, double-checking AI output, and limitations like prompt-dependence and vendor functional gaps.
Get email alerts on the Audit Technology topic
No spam. Unsubscribe anytime.
The internal audit division described its evolving use of data analytics and selective AI tools to improve coverage and productivity. Principal internal auditor Natalie Valdivia traced the program from ACL to Arbutus and said the division now uses continuous-audit scripts, Power BI dashboards, and AI tools for drafting and script corrections.
"We use AI to perform preliminary research and to assist with data commands... it makes our scripting and our formula creation much much faster," Valdivia said, but she stressed safeguards: documents are manually sanitized before being placed in external AI tools, and every AI-generated script, number and summary is double-checked by staff to protect accuracy and maintain auditor independence.
Valdivia noted that some government-licensed tools (Microsoft Copilot) lack specific features such as direct YouTube-to-minutes translation, so the division uses a combination of platforms for different strengths and continues to investigate enterprise audit-specific solutions. Committee members asked staff to research off-the-shelf enterprise audit AI products and to consider how ERP controls (Tyler Munis) interact with after-the-fact audit tools.
She emphasized that AI is an assistive tool, not a replacement for professional judgment: "We make sure that we manually verify every script, number, and summary drafted by AI." The division will continue using analytics to expand coverage while awaiting formal citywide AI policy and enterprise tool evaluations.

