Artificial Intelligence has firmly entered the boardroom. Companies are rethinking operating models, modernising infrastructure, and upskilling teams — all with the expectation that AI will deliver measurable business value. But as discussions at the Forbes AI Summit in Lviv showed, organisations are still grappling with a fundamental question: has AI truly moved beyond hype and into effective, scalable implementation?
I joined a panel titled “Next Big Things: Which areas of AI development will become critical for business in the near future?” Together with leaders from across industries — many outside traditional IT — we examined where AI is genuinely creating returns and where expectations still outpace reality.
One remark from a non-tech CEO captured a widespread frustration: AI vendors often promise transformational outcomes, yet after the systems are purchased and integrated, improvements fail to materialise. His company is now considering revenue-share models that pay not for effort, but for actual results.
This sentiment echoes a broader trend. MIT research suggests that up to 95% of AI projects fail. Why? In our discussion, two patterns stood out:
1. The “magic box” mentality
Because AI remains complex and abstract to many decision-makers, ideas often begin with assumptions like: “We’ll do this, we’ll add AI, and it AI will solve all our problems.” These approach pushes unrealistic ideas forward, companies spend time implementing them and later on figure out that IA simply is not capable to do all expected magic. Early expert involvement — especially in evaluating and stress-testing ideas — dramatically reduces this risk.
2. Overusing AI as a universal tool
When enthusiasm peaks, every problem begins to resemble an AI problem. However, AI is probabilistic, difficult to verify, and more costly to maintain compared to deterministic solutions. If an issue can be solved reliably without the aid of AI, it should be. Otherwise, organisations accumulate technical debt disguised as innovation.
I encounter both patterns frequently, and they remain among the biggest drivers of project failures.
Yet at the same time, there are organisations demonstrating what grounded, value-focused AI adoption looks like in practice.
A strong example is TecAlliance, where Christian Krause, Head of Data Science & Machine Learning, has systematically applied AI across products and processes. They automated technical documentation with generative models, optimised their hotline by analysing more than 12,000 voice messages each year, and used machine learning to detect and correct data anomalies — improving catalogue quality and reducing manual work. Each of these initiatives starts with a concrete operational need, a clear success metric, and a well-prepared digital environment. This is precisely the opposite of the “magic box” mentality: rigor, not hype, drives the results.
Another critical topic at the Summit was organisational readiness. Most companies aren’t building foundational models — they’re integrating tools created elsewhere. For them, the main bottleneck isn’t access to technology, but rather a lack of digital maturity.
I often describe it this way: imagine viewing your organisation through glasses that reveal only digital information – that is how AI will see your company. For many businesses, large parts of their processes, know-how, and decision-making frameworks simply disappear. Without a digital footprint, AI has nothing to learn from and no space to operate effectively.
The following 2-3 years, therefore, won’t be defined by who adopts the most AI tools, but by who builds the strongest digital foundations — accurate data, measurable processes – the more things have digital footprint the more visible the organization is for AI.
The Forbes AI Summit underscored a shift: the conversation is moving away from fascination with algorithms toward a more pragmatic focus on organisational design, measurable outcomes, and sustainable implementation. AI’s real promise lies not in novelty but in transforming how companies think, operate, and create value.
For leaders, the challenge is now clear: cut through the hype, invest in readiness, and build the conditions under which AI can deliver its full potential.
