Between algorithm and attitude: AI adoption in QA requires leadership, not technology

Why do AI projects fail in organizations? How do you deal with resistance to AI? These are just two of the many questions managers are asking themselves today.

More than half of organizations worldwide are actively investing in AI. Yet only 6% report a measurable impact on business results. The problem rarely lies in the technology itself, but more often in leadership.

It’s not a tooling problem

AI adoption in quality assurance (QA) is often framed as a tooling issue. However, the organizations that truly extract value from AI take a different approach: they treat AI as a leadership challenge. In QA environments, success is determined not by the technology itself, but by the quality of decision-making surrounding it.

The data supports this view. A study by McKinsey (2025) shows that 88% of organizations use AI in at least one business function, but that only one-third succeed in implementing it on a larger scale. The same study found that management attitude is the strongest predictor of success, more influential than tool selection or budget allocation. This brings us to the core issue: behavior.

Employees pay less attention to what managers say and more to what they do. A CTO who promotes AI but never uses an AI tool themselves is sending a clear message. Authentic leadership in an AI-driven environment means experimenting on your own, making your own mistakes, and making that process visible to others.

In practice, this translates into tangible and visible actions:

• Publicly share an AI experiment that did not work, helping to normalize experimentation and failure
• Ask teams about their AI experiences instead of only providing updates
• Recognize and reward early adopters. Not necessarily through bonuses, but through recognition and influence
• Create an internal channel where AI experiences can be shared openly and informally, without judgment or evaluation

There is, however, a common pitfall: positioning AI as a top-down decision that simply needs to be executed. This approach triggers exactly the resistance you are trying to avoid.

Managers who demonstrate genuine enthusiasm while also creating space for uncertainty foster a psychologically safe environment where adoption can grow organically. A culture of “Can I try this?” consistently outperforms a culture of “You must use this now.”

The Belgian testing landscape

What does this leadership challenge look like in a Belgian context?

The situation is nuanced. Many organizations have placed AI firmly on their strategic roadmap and are experimenting with applications in software testing; such as test automation and AI-generated test cases. However, broad adoption remains limited. QA teams face a constant battle: the pressure to test faster and more efficiently, set against a strong focus on reliability and risk mitigation. As a result, they tend to approach new technologies with caution.

A second tension further complicates matters. In many companies, there is a clear gap between management and employees. While C-level executives see the potential of AI, employees often experience a lack of clear guidelines, training, and psychological safety to experiment.

As a result, AI initiatives get stuck in the Proof of Concept phase instead of being systematically integrated into testing processes. Meanwhile, the costs of this delay continue to increase.

Shadow AI: the silent security risk

This gap has another consequence that many managers underestimate. The absence of strategic decisions and clear guidelines often leads to the rise of “shadow AI.” Employees turn to publicly available or personal AI accounts for professional purposes because no approved alternative exists.

The scale of this problem is greater than most managers realize. Menlo Security (2025) reported a 68% increase in the use of shadow AI, with 57% of users entering sensitive data into these tools. This is not a hypothetical risk, only a minority of organizations have policies in place to detect or manage shadow AI. A brief and practical AI integration plan, clearly defining which tools are available, for whom, and under what conditions, is the essential first step.

To learn more about shadow AI, read SQAI Suite’s in-depth blog: Shadow AI Is a Security Risk.

How can organizations successfully drive AI adoption?

The challenges are clear, but what does an effective approach look like? For AI implementation within QA teams to succeed, a people-centered approach is essential. It starts with recognizing that adoption requires behavioral change, not software/tool installation.

In practical terms, this means:

• Invest not only in technology but also in targeted support and enablement. Create safe environments where teams can explore AI without immediate performance pressure
• Define clear use cases that align with the daily reality of testers, making the value of AI tangible rather than abstract
• Identify internal ambassadors: testers who are intrinsically motivated to work with AI and can inspire others through their enthusiasm
• Address common concerns openly and transparently, such as “Will AI replace my team?” or “Can we trust the results?” Present AI as a complement to human expertise, not as a replacement
• Build confidence through small, visible successes. One successful use case is often more convincing than an ambitious roadmap full of promises
• Establish a clear AI policy for QA teams, specifying which tools are approved, who may use them, and under what conditions. This helps prevent shadow AI while creating the psychological safety needed for experimentation
• Invest in an objective assessment (such as our Brightscan) to determine the current level of AI maturity within your QA teams, identify where AI can have the greatest impact, and define how and when it should be integrated into your existing toolset.

Organizations that succeed in these areas achieve not only faster adoption but also more sustainable value creation.

Successful AI adoption starts with leadership

The technology is in place, the resources are available, and the business case has already been drawn up. What most QA organizations are missing is not another AI platform, it is leadership that takes the human side of adoption seriously.

Emotional intelligence enables leaders to manage resistance and uncertainty while making sound ethical decisions. It allows them to guide AI initiatives effectively without compromising the reliability and quality standards that QA demands. The question is therefore not whether your organization will implement AI. The real question is whether management is prepared to lead that adoption.

Sources

1. McKinsey & Company. (2025). Superagency in the Workplace: Empowering People to Unlock AI’s Full Potential at Work. McKinsey.
2. Menlo Security. (2025). 2025 Report: How AI Is Shaping the Modern Workspace. Menlo Security.