AI Implementation: AI Initiatives Fall Short of Expectations

While it’s hard to find anyone who isn’t excited about AI’s potential to enhance productivity, a recent survey from the Computer Technology Industry Association (CompTIA) of more than 1,100 US business respondents says that, so far, reality has fallen far short of expectations. Here’s an excerpt from CompTIA’s report on the survey’s findings:

A majority of companies (79%) in this research report backtracking from an AI solution that did not work out as expected when attempting to substitute AI for human work. This may entail relatively minor backtracking, such as a scenario whereby a marketing team tests a gen AI tool to assist in generating content for the company’s blog. They may find some of the output of an acceptable quality, while other portions of it may not meet editorial standards. The marketing team may backtrack to reassess finding the right AI tool for the job, along with user training to finetune prompts, and a process playbook to operationalize output reviews.

More significant backtracks from AI deployments may entail scenarios involving security risks, legal exposure, compromised intellectual property, or lost revenue due to a poor customer experience.

The report says that performance and technical challenges are the most common reasons that companies backtrack from an AI deployment, costs also enter into the equation.  An AI solution may work fine, but the ROI may not justify the costs associated with it.