The old saying “you are what you eat” can also be applied to your Artificial Intelligence (“AI”) ERP project development and implementation. While AI is being touted as the “innovation of the century,” it can only be as functional as the quality of data that you are feeding into it. Will your AI ERP integration be robust and healthy? Or will it malfunction, show weaknesses, and ultimately be rendered useless? Having the right inputs as well as the right people deciding about those inputs (much like a nutritionist might help someone learn about healthy eating) is imperative.
However, unlike engaging with a nutritionist who is limited to giving you dietary advice, your company is likely to retain an ERP vendor that can and should take the lion’s share of responsibility and liability for integrating AI. Because inserting AI into your ERP system too soon could have disastrous effects on your company’s bottom line, it is incumbent on your company to have a thorough legal review of the controlling agreements to carefully delineate threshold tasks and establish clear lines of accountability.
While there are endless considerations that your legal counsel should ensure are in a vendor contract, when AI is in play, we urge your company to specifically build in contractual provisions that ensure data integration preparedness and your vendor’s specialized expertise.
It should come as no surprise that AI needs organized and consistent data to be effective. It needs to be clean and, ideally, stored in a central database. That said, it is equally essential that a company’s data be integrated across all systems and departments, as well. Data silos are a common hang-up in the data integration process.
Data silos are isolated collections of data that prevent data sharing between different departments, systems and business units. When data becomes siloed, organizations can struggle to maintain data quality and make data-driven decisions.
As a result of data silos, teams often end up working with outdated, fragmented or inconsistent data. Data quality degrades, and operational inefficiencies arise from duplicated workflows and redundant data storage. Big data, machine learning (ML) and artificial intelligence (AI) initiatives can all suffer.
Bottom Line: Without quality data, companies will struggle withinaccuracy, faulty predictions, and potentially a lack of trust and buy-in. Spend some time on the front end fixing and streamlining data and ensuring that historical data is accurate. Conduct a data audit as ‘step one’ in any AI ERP project.
A lack of expertise in data science and AI integration is a common issue within many companies. Organizations will either need to invest time and resources into training their teams or hire experts from outside to assist with the process. As noted in a The Future of ERP blog post:
AI and machine learning require specialized skills that many businesses—especially small ones—don’t currently have in-house.
Further, even the vendors that do have the necessary expertise haven’t necessarily had the time to build a deep bench of such experts. Nothing is more demoralizing than meeting with an impressive “A Team” only to find yourself trusting your company’s future with a less skilled and experienced “B Team.”
Bottom Line: With an AI ERP project investment only as valuable as it is useful, having skilled people to manage AI models is a necessity. Consider the following:
* * *
Bad data and lack of expertise can do more than just derail your company’s goals. It can lead to real financial losses as well as catastrophic legal consequences.
While no risks can be entirely evaded, they can be managed – especially at the front end. It is imperative that you negotiate terms and conditions which protect your interests in any AI contracts, and, as always, we are here to help!
Published on September 4, 2025
Software licensors are known for vague contracts—they’ve made a business of it.
Read the latest industry news.
