How CSOs at Synchrony, Penske, DTEX Tie AI Bets to Business Lines

Maran Nalluswami of Synchrony described how the consumer lender uses machine learning to personalize credit offers and predict delinquencies. Synchrony holds $40 billion in credit card receivables. A one-basis-point improvement in charge-off rates is material. Nalluswami stressed that every model must be explainable to regulators. Human underwriters stay in the loop on any decision that changes a credit line. The bank tests new models against current approval and loss rates before scaling.

The three executives shared a skepticism about broad AI transformations. Each company filters ideas through a specific ROI gate. Penske cuts projects that do not solve a named customer issue. DTEX requires a time saving per analyst that can be measured. Synchrony benchmarks against current loss rates. None of the three CSOs described a moonshot.

Payoffs will show up in unit economics rather than a single revenue jump. Synchrony could see lower charge-off rates over the next two to three quarters. Penske's routing efficiency might reduce operating costs as a percentage of revenue. DTEX's improved alert accuracy can support higher subscription pricing. None of the executives gave a timeline for when the spending would hit the income statement as a measurable tailwind. All three said their companies would report progress on the next earnings call.