My colleague, Isabelle Papoulias, published a very thoughtful post on LinkedIn and I had to share it here. She talks about the fact that with the potential for increases in productivity and other benefits of AI technology, many folks are not talking about a critical component of a successful implementation: People.  

I completely agree with Papoulias that AI is a change management challenge. Employees don’t like to change the way they do things. They need to understand the vision and the benefits. They need a compelling “why” to make changes. 

A good comparison might be ERP system implementations because they too touch the whole organization. When ERP implementations fail, it usually isn’t the technology that was the problem – it was the people. A critical piece of the project needed to be change management and training. Oddly enough, when I was helping to write proposals at a big consulting firm, this was the part of the project that the prospective client thought was fluff and wanted to cut. 

It was not fluff. It was the key to success. 

Related is the fact that in some race to be first that I don’t quite understand, senior management at many organizations are pushing these tools down to employees, without a defined strategy with goals, and mandating employees use the tools, often without proper training and guidelines. 

Even one of the leaders in the industry had a big oops recently. On March 31, 2026, AI company Anthropic accidentally leaked over 500,000 lines of proprietary source code for its Claude Code tool via a public software update, resulting in a viral, widespread dissemination of its intellectual property across social media. (Source) 

Please note: This type of IP risk applies to many companies and proper training, guidelines, and governance is needed. 

I am so glad Papoulias also mentioned the real fear that many workers have of being replaced. Will workers be motivated to leverage a technology that may take over their job at some point? 

My opinion, and I am hearing this more frequently from others, is most organizations are not taking the time they need on the front end to think through the components of an AI implementation and will end up having a failed or less successful implementation because they didn’t do the required planning initially. It will likely be slower and more expensive to go back and do it later. 

You can scroll down to read the whole post. 

Photo by Alex Kotliarskyi on Unsplash