Trust, not budget: Why AI adoption in finance comes down to governance
SiliconANGLE Ryan Stevens ● Covered by 7 sources
AI is creeping into finance through the tools teams already use. The hard part isn’t budget; it’s trusting the answers before they hit filings.
Based on reporting by SiliconANGLE, Ryan Stevens — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
AI is moving into corporate finance from the systems accountants already live in: software, reporting tools and ERP platforms. That’s why the real bottleneck isn’t money. It’s governance. If the output can end up in a filing that shareholders and regulators read, finance teams need a way to prove the machine didn’t just sound confident — it was right.
Christie Kozlik, chief accounting officer at Accel Entertainment, framed it bluntly during an interview at Amplify with theCUBE’s Krista Case and Alison Kosik. “It’s one thing to turn on Claude. It’s another thing to use it smart,” she said. That distinction matters more in regulated businesses, where trust is not a nice-to-have. It’s the whole job.
Accel’s scale makes the point concrete. The company operates 29,000 slot machines in 4,700 locations across 10 states, with data coming in from dozens of source systems. In that kind of environment, a model that answers in microseconds can be a help or a headache, because it can sound equally certain either way. Kozlik said the old human cues are gone. No body language. No tone. Just confidence.
So her team is treating AI like any other controlled process. Define the input. Define the expected output. Document the review. Bring internal and external auditors in early, not after the fact. She also wants ROI measured in clear operational terms, like cutting the financial close from seven days to five. Her near-term focus is SEC reporting and the company’s ERP buildout, where the data lineage is already known and the audit trail is already part of the game.
Kozlik’s suggested filter is almost embarrassingly practical: rate each footnote in a 10-K or 10-Q and decide whether it is AI bot-ready. That’s the right instinct. Finance doesn’t need another demo that dazzles for five minutes and then wanders into a filing with a straight face.
My take — AI-written commentary, not fact-checked reporting
This is the grown-up version of AI adoption, and it’s overdue. Finance won’t be won by louder demos or bigger budgets; it’ll be won by the boring stuff everyone loves to skip: inputs, outputs, review, audit trails. The hype crowd hates that answer because it doesn’t fit on a slide, which is exactly why it’s the right one.
Read more about this at: SiliconANGLE
