Share with your CISO
Financial services firms rushing to deploy AI are hitting a wall they didn’t budget for: governance. At FINRA’s 2026 Advertising Regulation Conference, compliance leaders from Wells Fargo and peer institutions warned that AI governance integration consistently exceeds what firms anticipate in time, headcount, and process redesign. Wells Fargo’s Stephen Bard put it plainly: “Don’t underestimate the governance work involved.” The article is paywalled beyond the opening exchange, but the signal from regulated-industry practitioners is clear enough to act on.
What this means for your business
Based on limited public information from Compliance Week, the pattern here is well-established even if the specifics are behind a paywall. Firms in regulated industries, financial services especially, tend to treat AI governance as a compliance checkbox they’ll handle after deployment, then discover it’s load-bearing infrastructure that should have been poured first. If your organization is already running AI tools in customer-facing or regulated workflows and your governance framework is still a draft policy document, you’re in the cohort this warning is aimed at.
The integration problem Bard names isn’t primarily a technology problem. It’s an organizational one. Supervisory controls, the processes that ensure a human reviewer can audit and explain an AI-assisted decision, don’t map cleanly onto existing approval chains. Building them requires compliance, legal, IT, and business-line leaders to agree on accountability at a granular level most firms have never attempted. That work takes longer than any vendor’s implementation timeline suggests, and it can’t be parallelized indefinitely by throwing headcount at it.
The firms that come out of this period in the stronger position won’t be the fastest deployers. They’ll be the ones whose governance scaffolding was built to survive a regulator’s examination, not just an internal audit. FINRA convening this conversation in 2026 suggests the exam pressure is real and near-term. If you’d revise this read, the falsification condition is simple: show me a major financial services firm that deployed AI at scale without pre-built governance and didn’t face a material supervisory finding. Until that case exists, the compliance leaders at this panel are right.
Based on reporting from AI governance work is harder than firms expect, compliance leaders warn, originally published 2026-11-09 18:17:00.

