Agentic AI Is Testing the Limits of Legacy ERP
Agentic AI Is Putting Legacy ERP Systems to the Test
Agentic AI (artificial intelligence) is changing what businesses expect from their ERP (enterprise resource planning) systems, but according to Jack M. Germain in his CRM Buyer article, organizations running decades-old ERP environments may discover that adding AI doesn’t necessarily simplify operations—it can expose the complexity already hiding underneath. AI agents can automate workflows, make recommendations, and potentially take action across business systems. But to do that effectively, they need reliable, consistent data. When customer, inventory, order, or financial information is spread across multiple systems, agents can end up spending more time finding and reconciling information than actually making decisions.
The article points to several warning signs that an ERP environment may be becoming more complicated than valuable: rising integration costs, customizations that require workarounds or spreadsheets, and reports that depend on reconciling data from multiple systems. These problems become even more significant when AI agents need to interact with those same systems. That raises an important question for CIOs: Where does the authoritative version of the business data actually live? As AI agents move from simply providing recommendations to taking action, organizations need clear systems of record, defined permissions, and centralized audit trails. The good news is that companies don’t necessarily need to throw away their existing ERP investment. The article suggests starting with individual workflows where the current system creates friction—such as scheduling, shipping, or quoting—and introducing newer capabilities alongside the ERP. Future development can then gradually shift toward a more modern architecture.
The biggest takeaway is that AI doesn’t eliminate ERP complexity—it can amplify it. If an organization wants AI agents to make faster, smarter decisions, it first needs to give them clean, consistent data and a clear foundation to work from.
Retiring Lawson, PeopleSoft, or Oracle? APIX archives the entire application — every table, every year, attachments and security included — into your own AWS account in about 30 days, so you can decommission the legacy system and keep full access to the history.
