The Compliance Layer Dealers Forgot to Build Before Adding AI

The conversation captured on CBT News' Training Camp, featuring Jim Ganther of Mosaic Compliance Services, arrives at a moment when dealership technology adoption has comfortably outpaced dealership governance. Ganther's core observation, drawn from two decades advising dealers, is that compliance exposure no longer sits neatly within advertising and finance departments. It now extends to every text message, social post and AI-generated recommendation an employee produces on the dealership's behalf. That expansion of liability is not new in principle, but AI accelerates it in ways that most dealer groups have not yet priced into their risk models.

Ganther's specific warning about AI in lending decisions, hiring and advertising deserves attention from leadership teams beyond the compliance function. These are precisely the areas where discrimination and disclosure law intersect with automated recommendation engines, and where a dealership could find itself liable for a decision no human meaningfully reviewed. His prescribed remedy is unglamorous but sound: treat AI as a research assistant whose output is verified before use, rather than a decision-maker whose conclusions are simply implemented. He draws a direct parallel to his own legal practice, where AI has sped up research substantially while still requiring attorneys to check every result. For dealers, the practical translation is that AI can compress the time spent drafting an ad or screening a lending file, but it cannot yet compress the accountability that follows if that output turns out to be wrong.

There is a structural point buried in Ganther's remarks that matters as much as the AI warning itself. He describes the compliance industry moving from narrow specialists in F&I, auditing or data security towards full-service platforms covering variable and fixed operations, auditing and cybersecurity under one vendor relationship. For dealer principals juggling an already crowded technology stack, that consolidation is a rational response to complexity rather than a sales pitch, and it suggests procurement teams should be asking whether their compliance vendors can actually keep pace with AI-specific risk, or whether they are still organised around pre-AI categories of exposure.

Ganther closes on Mosaic's long-standing motto, making virtue profitable, framing compliance not as a cost centre but as a condition for durable financial performance. That framing is easy to dismiss as marketing until one considers the asymmetry involved: the upside of AI-assisted efficiency is incremental, while the downside of an unreviewed AI decision in lending or hiring can be existential for a dealer group's reputation and balance sheet. The leaders who treat oversight as infrastructure, built alongside AI deployment rather than bolted on afterwards, are the ones likely to capture the efficiency gains without inheriting the liability that comes free with them.