Good management practice taught us for a long time that better decisions depend on more information. So we invested in dashboards, reports, indicators. Yet leadership confidence is falling. In PwC's 29th global survey (2026), the share of executives confident in their own company's growth dropped to 30%, the lowest level in four years, down from 56% in 2022.
The question that occupies these leaders most isn't about the market or the competition. It's whether they're moving fast enough to keep up with technology, especially artificial intelligence (AI). Despite the investment, 65% report little or no revenue change from adopting AI (67% in Portugal). As a result, most remain stuck in pilots that never reach scale.
The problem isn't a lack of data, or even a lack of technology. It's the gap between having information and turning it into a timely decision. That's where the real competitive advantage sits.
A slow decision is an opportunity that goes to someone else. A wrong decision carries a cost measured in years. A reactive decision, made once the problem is already at the door, is almost always a disadvantage. But the weakness isn't in the intelligence of the person deciding. It's in deciding late, with the wrong information, or without prior analysis of the external context.
We know the classic stories of Nokia, Kodak, or Blockbuster as cases of failure to anticipate. But the examples keep coming. In 2025, NVIDIA posted the largest single-day stock drop in history, around $600 billion, in reaction to the launch of DeepSeek, a Chinese AI model that showed competitive AI could be trained with a fraction of the resources thought necessary.
That episode didn't stand alone. A few days ago, an open model from another Chinese company, Z.ai, was rated by security researchers as on par with the best American models, at roughly half the cost. At the same time, advanced models like Anthropic's Mythos/Fable were held back by export controls and safety reviews. The advantage once considered secure has narrowed sharply.
The signals had been there for a long time: export restrictions on chips to China, Chinese investment in AI. What the market lacked wasn't information, it was anticipation. Which raises the question: if a disruptive innovation is born outside an organization's radar, how long does it take to reach us? In a world where the answer is measured in weeks, how many innovations are growing right now outside the field of view of those making decisions?
The same weakness shows up, more quietly but just as significantly, in how companies handle European regulation. The Draghi report on European competitiveness counted around 13,000 legal acts approved by the European Union between 2019 and 2024. The problem is the volume, and above all, how this legislation lands inside organizations.
Many companies treated GDPR as a standalone project, with years of system and procedure adjustments as a result. Now NIS2 has arrived, the cybersecurity directive that can carry fines up to €10 million or 2% of turnover, with personal and non-delegable liability for management bodies. In financial services, DORA has applied since January 2025. And European regulation itself keeps changing. The Digital Omnibus package, the simplification proposal presented by the European Commission in 2025, amends GDPR, NIS2, DORA, and the AI Act itself, legislation that in several cases hasn't even had time to fully apply yet.
Many organizations still work in regulatory silos. Legal looks at GDPR, IT looks at NIS2, finance looks at DORA. It's precisely in joining up this regulatory governance that the risk, or the opportunity, lies for those who prepare in advance.
This is an uncomfortable lesson, since the world isn't returning to business as usual. As the World Economic Forum notes, volatility has become the norm, not the exception.
This is where technology comes in, but it's essential to understand its limits. AI and data tools don't make strategic decisions for us, nor should they. What they do is shorten the distance between the signal and the decision: spotting patterns early, cross-referencing sources that used to sit apart, testing scenarios, and freeing up time for better-grounded human analysis.
Unlike traditional market research methods, limited by cost in their means and sources, this new model changes the pattern. With proper data curation, the combination of AI-driven research and continuous use of many credible sources can bring a major competitive advantage in reading the right signals and deciding with quality and speed. Even so, technology without method, governance, and someone who owns the decision adds nothing but noise.
Anticipating, then, isn't about predicting the future. It's about building the capacity to read signals and decide before being forced to. Companies and institutions that treat this capacity as a discipline, not an annual report, end up on the right side of the next disruption.
Eisenhower used to say that "plans are useless, but planning is indispensable." What looks like a paradox isn't one: the value isn't in a fixed plan, it's in the discipline of understanding what needs adjusting. That commitment is what separates those who anticipate from those who already know everything that went wrong.
Originally published in Link to Leaders, July 27, 2026.