top of page

Stop shadow AI with the easy button

Last week I shared dinner with a group of fellow senior technology & business consultants. We were brought together by a common “1st degree” connection, with whom we had all worked before. As we shared experiences, we all mentioned how @Eric Norden and his team(s) had all been “clutch” and laser focused on contributing to the success of their then current business and mission, rather than a being a “trouble ticket management crew”. Refreshing, I know.


We inevitably got to the topic of AI (shocker), and how businesses are trying to deploy the technology for either internal use, or incorporating it to their products and services offerings. The general consensus at the table was along the lines of “everyone wants AI, but  most don’t even know where and how to start”.


I mentioned that shadow AI is spreading like a wildfire in the SMB world. Department heads and senior leaders are urging “early adopters” within their businesses to use personal or free AI applications, and research how they may be used in running their businesses. Without skipping a beat, @Alan Kisling said about the existence of shadow AI [or other technology] efforts: “it is because working with their [internal team] is a PITA.”


No need to wonder then where and how to start deploying AI.


The thinker ponders on Shadow AI
Wondering how to stop Shadow AI?

If your business or business unit is thinking about using AI to improve business performance, here is the easy button: work with a technology deployment team that has breadth of business experience, practical technical knowledge, and who truly enjoys trying “the new stuff”.


The “easy” button has


Business experience; deep, wide, and in abundance.

Deploying AI capabilities within your existing technical stack is more than making every available AI app feature and function available to the team. That is both easy and sloppy. Quite frankly, that may be the fastest way to thwart AI adoption, and to keep your corporate legal team booked solid for the next year.


Internal, external, or combination thereof – start with a team that has hands on experience in running a business similar to yours. A team that has both individual and collective experience at every level of performance, and across departments.


AI [any technology] implementation should be thoughtfully considered, so that there is positive, clear and measurable impact on the business, with little to no burden on user experience. Knowing what information is needed, when, where, by who, and how it should be delivered, are dimensions of the problem that have little to do with apps features and functions.

These implementation questions are second nature to, and are best solved by the individuals and the collective hive that has done the job, and worked across functions.


The AI technology capabilities that your business plan calls for are out there. A team with deep and wide business operations experience will make easy work of integrating not only the AI apps to the current tech stack, but also and more importantly to the business’ workflows.



Technical knowledge and integration creativity

Beyond business experience, the technology deployment team has to channel some “MacGyver” energy to the next level.


For all of the capabilities that AI applications have, universal integration is not one of them. They are largely islands of functionality that are single purpose driven, with disparate interfaces and prompt engineering requirements.


Back to just making every available AI app and feature available to the team… it will thwart adoption for sure.


Make sure that you work with a technology deployment team that simplifies prompt engineering, such that no user is required to learn multiple tech dialects while they are trying to do their job.


More importantly, the use of AI is multi-modal, and with single or narrow focus applications, users need to work with many of them to do one job. This is where technical knowledge and creativity comes in. Beyond API calls and “Zaps”; technical creativity and knowledge is required to modify the prompts to, and the returns from the AI applications so that the “post AI production” workload for the user is reduced. The whole point of using AI was to increase efficiency and effectiveness. Giving three more AI tools and twenty additional workflow steps to the users defeats that purpose.



A joyful “unboxing” spirit

Personality and positive mental attitude are often overlooked or shadowed by technical credentials and business experience. The team has to be good at, but also enjoy what they do.


Injecting new technology like AI to the technical stack is no joke. The workload can be overwhelming, especially when IT’s manpower budget is constricted to the max. “Please throw more work at me” said no one ever who works in IT.


Make sure you have a technology deployment team that is not going to be overburdened by doing their job, a new AI deployment project, and also briefing leadership on why they are behind on both.


In order to make the implementation of new technology bearable, productive and successful, seek a technology implementation team that welcomes the challenges of testing new technology as they incorporate it to the business. Select a team of pioneer spirits that welcomes and loves uncertainty, and will not be deterred by the inevitable failures and challenges that will test the project plan.





Shadow AI has no place in your business. It is an approach that eventually catches with the business and its leaders, and not in a good way.


If you want to deploy AI in your business correctly, work with a team that has breadth of business experience, practical technical knowledge, and who truly enjoys trying “the new stuff”. Deploy AI [and other IT solutions] starting your journey with a team that is an easy button, and not a PITA.


12 views0 comments

Recent Posts

See All


bottom of page