Kozlik: AI in finance gated by trust, not budget

TheCUBE's Krista Case and co-host Alison Kosik spoke with Kozlik at Amplify, in a segment sponsored by Workiva. Kozlik's starting point is that AI is already arriving from the inside out, through the software, reporting platforms and ERP systems that accounting teams already run, so the practical question is not whether to adopt it but how to check what it produces.

Her discipline is deliberately old-fashioned: every process gets a defined input, an expected output and a documented review, with internal and external auditors brought to the table early rather than after the fact. She frames the work as risk rating the data itself, footnote by footnote, rather than treating governance as a single blanket policy.

The stakes hinge on whether that review discipline can keep pace as AI moves into SEC reporting and the ERP buildout, the heavily audited areas she has chosen first. If the checks hold there, finance teams in regulated industries may have a template for moving faster; if they do not, the confidence of the model's answers is likely to be the thing that slows adoption.