One fun fact about me is that I wrote my first Master’s Thesis on shadow economy. I’ve always found it fascinating how many facets this phenomenon has worldwide. Different countries opt for different approaches, so that same activity can become illegal by virtue of crossing a border (and within the US, even a border between two states). However, forbidding and not regulating certain economic activities doesn’t ban them from existence. This applies to both the shadow economy and the usage of AI technologies. If anything, restrictions only make the “forbidden fruit” more tempting. Moreover, it means that employees use the technology secretly, finding creative ways to smuggle outputs back into the work environment, sometimes exchanging meaningful glances at meetings and creating a subculture of their own around it.

The mesmerizing beauty of such dark phenomena as shadow economy lies in the chain reaction potential they embody. Let’s look closer at those chains...

Inaccurate Economic Data

It always starts with basics, and data is the most basic and crucial of them all. The more takes place under the radar, the more inaccurate the picture of reality – including current workload – is going to be. Skewed understanding of the workload impacts capacity planning, setting off a domino effect on production planning and general management. As the actual economic data and your understanding of it drift further apart, you start seeing how a simple yes-or-no question could be answered with “technically no, practically yes” in the Soviet Union. How is this going to impact decision-making?

Challenges in Resource Allocation & Development

One consequence of this drift is a gradual loss of the ability to realistically assess which areas require improvement and development. This makes it difficult to know where to apply your most precious resource: attention. As your energy flows where your attention goes, and money follows, it’s only pure luck if you manage to make the right development decision under such circumstances. What implications does this have for internal dynamics?

Distortion and Unfair Competition

It’s no accident that I mentioned meaningful glances at meetings. Just like in the context of shadow economy, sharing a secret becomes a true bonding experience for colleagues involved in it, more effective than any team-building event you were planning. This creates an opportunity for two poles within your organization to emerge—those who “get it” and those who don’t. This introduces an “us-them” complexity into organizational dynamics that you don’t necessarily want, as one group feels unfairly treated and disadvantaged while the other is on the verge of “ruling the world.” From this point on, you can start the countdown to the next War of Roses – on your watch. What does that mean for you?

Undermining of Social Contract and Trust in Management

The fact that these dynamics take place undermines the implicit social contract between you and your employees. Even more crucially, it undermines your authority and the trust of your employees that you have an accurate picture of reality and possess the ability to manage it and truly drive the organization. How is this going to impact your organization’s long-term trajectory?

Slippery Slope & Potential for Increase in Crime

Involvement in shadow economy does bring some short-term benefit, but it’s like living on credit. It speeds up certain processes, helps move forward in the absence of regulation, etc. However, the price is high and due in the mid- to long-term. With less trust in management on the one side and the knowledge that one is already doing something forbidden on the other, the border between acceptable and unacceptable behaviors blurs. This results in a steady erosion of values and principles, undermining the very foundation of an organization.

As you can see, shadow AI impacts everything – from short-term data accuracy to long-term values and the foundation of an organization. The ultimate outcome is a suboptimal equilibrium for an organization, in the worst case with a set of institutionalized routines supporting this self-propagating mechanism: more vulnerability, exposure, lack of control, and an ever-growing shadow.

This downward spiral isn’t exactly uplifting, is it? So, let’s change direction! If the millennial history of shadow economy regulation teaches us one thing, it’s this: bring the matter into the light as soon as possible! Actively managing AI technology puts you in a position of strength. You’ll be able to stay on top of things, integrate new technology into your processes, and enable your organization to reap its benefits. Now, here’s the good news, and it’s coming from an unexpected direction: global legal initiatives like the AI Act. Far from being just a compliance hurdle, they’re a blessing in disguise. In addition to providing a framework they also create a strong impetus for compliance, giving you an opportunity to speed up integration.

How exactly do you go about it? That’s an excellent question, and we’ll explore it in the next article, two weeks from now – same time, same place.

Until then, let me leave you with this piece of advice: don’t try to apply an IT solution for an organizational and cultural issue…