Sarah Mattalian: Why is having the correct ERP foundation important? What happens if manufacturers try to scale AI without the right ERP foundation? 

Nancy Majure: I've been saying for years, we've been saying for years, ERP is not going to go away. ERP is that foundation. It's that system of record. It is that single version of the truth. It's that place where your production data, your quality specifications, and your financials, they all agree. They're all reconciled together. I think it's important to note that AI absolutely will not fix a shaky foundation. 

In fact, it will amplify it. At machine speed, it will amplify it. That ERP is that foundational, deterministic system that really puts those guardrails in place to ensure that you're not putting your operations at risk. 

See also: Podcast: How AI is altering the landscape of manufacturing cybersecurity 

The most common pitfall is essentially I see manufacturers bolting these kind of generalist third-party AI systems on top of their ERP instead of building into it. First of all, they're not thinking about the importance of governance. ERP by its very nature is intended to be deterministic. So always, ‘I put in A, I'm going to always get out B.’ There have been decades of work that has gone into building out the business logic and the governance and the controls and the rules around what happens within your business.  

So, since ERP is deterministic, a lot of generalist AI is going to provide you those probabilistic responses. It's going to give you probably the most likely answer as opposed to what would be the right answer. When that AI is sitting outside of your governed environment, it doesn't take much. A single bad signal and hallucinated potential spike in demand that can trigger a snowball effect that can actually be very damaging to a business, triggering over purchasing, and triggering overstock. That first pitfall is trying to think that you can get outside of the boundaries of an ERP deterministic system.  

See also: AI in manufacturing won’t deliver until supplier ecosystems are connected 

But the second [pitfall], and I think it's all too common, especially manufacturers that have been around for a while, is fragmentation. You'll see it so often in companies, especially where they don't have a solid ERP, a single version of the truth. You have production data in one system, quality in another, maintenance in another, perhaps you have your financials someplace else. And the reality of their current life without AI is that they spend a lot of time and a lot of effort in reconciliation between those systems.

They do not always agree. And what happens is if you then try to scale AI on top of all of that, it's literally just going to scale the disagreements between the systems.  

SM: What are some preventative measures that manufacturers can take so they can have the right ERP foundation?  

NM: You know, it doesn't have to be rocket science. I think that the very first and most important thing that a manufacturer can do—because all manufacturers are feeling the pressure, they have to adopt this—but very pragmatically, start by mapping the AI you already have running.  

See also: Black Kite: Ransomware increasing across all metrics in 2026 

Now, I want to be sure to mention this. That includes the official AI as well as the unofficial AI, because you just simply can't govern what you can't see. I see it on a regular basis when I encourage companies to take a step back, map out what they have, they're always surprised. by how much shadow AI employees have introduced.  

It's not a nefarious scheme on the fact that the employees, employees want to do their job. They want to do it well, they want to do it efficiently, and they want to move faster. And so that's really the first piece. Understand what you have and understand your exposure.