Your Core Vendor Is Already Selling You AI. Do You Know What You Bought?
Two of the biggest core and payments vendors serving credit unions just wrapped up earnings season, and the numbers tell a consistent story: AI isn't a side project anymore, it's a selling point baked into the core contract.
Jack Henry closed fiscal 2026 with 58 competitive core wins, its best year in more than two decades, and reported that AI-enabled products and advanced treasury tools are becoming a larger share of total revenue. The company now has 22 AI-enabled products live, with more than 20 additional capabilities planned in the next six months, covering things like financial crime detection and multilingual support. FIS told a similar story: more than 200 customers are live on its AI products, with over 500 more opportunities in its sales pipeline.
That's the good news version. The more useful question for a credit union isn't whether your vendor is selling AI, it's what "AI" means inside the specific bundle you're being pitched.
What's usually really in the bundle
Based on what these vendors are actually shipping, the AI capabilities landing in production tend to cluster around a few defined jobs: fraud and financial-crime detection, document retrieval and processing, and internal workflow automation that speeds up staff, not member-facing decisioning. That's meaningfully different from an AI system that's making or heavily influencing a lending or account decision. Both get marketed as "AI-enabled," but they carry very different governance obligations.
Jack Henry's own numbers are a useful tell here. The company reported using AI internally to increase engineering throughput, cut software defects by roughly 30%, and reduce manual ticket volume by around 70%. Those are real, measurable gains, and they're mostly happening behind the scenes, in how the vendor builds and supports its own software, not necessarily in a tool that's scoring your members.
Three questions worth asking before you sign
When a rep says a renewal or upgrade "includes AI," it's worth getting specific before you agree to anything: Is this feature making a decision about a member, or supporting a staff member who makes the decision? What data trained it, and is any of that your credit union's own member data? And if an examiner asked how the feature was validated, what would the vendor actually be able to hand over?
None of this means walk away from AI-enabled core features. The adoption numbers suggest this is where core banking technology is headed regardless. It means treating "AI-enabled" as a starting point for due diligence, not a finish line. A feature that speeds up document retrieval for your staff carries a very different risk profile than one influencing whether a member gets approved for a loan, and your vendor contract and internal documentation should reflect that difference.
Sources