The void that AI does not fill: leadership and alignment in Costa Rican nearshoring
The problem is not AI. The problem is the lack of judgment with which it is being adopted.
At the beginning of the year, I facilitated an alignment exercise during a leadership summit for a multinational corporation in Costa Rica. Around the table were thirteen leaders with years of experience, many with direct reports in three or four countries. The dynamic was simple: map out where the team is heading in the next two years and how that direction is communicated downwards.
The exercise stalled on the second question. One manager voiced it aloud, with an honesty I appreciated: they themselves did not know where the company was headed in the long term. And if they didn't know, they had no way of providing security to their teams. The silence that followed was the most relevant data point of the day.
The first-quarter numbers are clear. According to a report by El Financiero, globally over 80,000 people lost their jobs in technology companies in the first three months of 2026. Oracle laid off between 25,000 and 30,000 employees, about 18 percent of its workforce. Meta is preparing a cut that could reach 20 percent of its staff, while projecting to invest 135 billion dollars in AI infrastructure this year alone.
What is unsettling is not the layoffs. It is the reason. These companies are not cutting because they are doing poorly. Oracle was breaking revenue records. Mark Zuckerberg put it plainly: projects that previously required large teams can now be done by a single, highly specialized individual.
Costa Rica is a direct nearshore location for these companies: the multinationals operating from here report to and respond to the pace of their parent companies in the United States. Decisions made in Austin, Palo Alto, or Seattle land in Heredia, Escazú, and Belén within a few weeks. CINDE's data confirms this: the multinational sector supports 196,310 jobs in the country, but in 2025 it grew by only 3,259 positions, far below previous years. Corporate services, a historical engine of specialized employment, contracted for the first time in nearly 2,000 positions.
In conversations with leaders of global banks in recent months, I have encountered a recurring phrase: meeting and exceeding goals is no longer enough; everything is at stake. When that phrase trickles down the chain of command, it reaches the teams in Costa Rica. And there begins the question that few dare to ask aloud: are we building more efficient companies, or simply more fragile ones?
The problem is not AI. The problem is the lack of judgment with which it is being adopted.
In this year's board meetings, a silent decision is being made: replace or empower. Replacing means cutting staff and filling the gap with an inexpensive tool. Empowering means keeping the team that already knows the business and giving them AI to do more and better. The first path appears more efficient in the quarter. The second builds a company that remains alive in three years.
I say this from experience. I use AI every day. Claude, Notion, GitHub, Lovable, Google Workspace. They are part of my operations as a consultant. They make me faster, more organized, more precise. But without my prompts, without my judgment, without my corrections, these tools do not know what my business needs. They do not understand why a client in Germany requires a different approach than one in Houston. I provide that layer.
That is the point being missed in many leadership rooms. AI is excellent when there is a person with judgment behind it. Without it, the tool functions halfway, and no one understands why. The error is not in adopting AI. It is in treating it as a technical decision when it is actually a leadership decision.
Quarterly metrics do not capture what is happening in the remaining teams. There is a question circulating in hallways, internal chats, and lunches: who will be next? It is not said in staff meetings, but it is there. And it changes behavior.
Employees enter survival mode. They do the minimum to avoid drawing attention. They do not propose risky ideas. Creativity, the very thing companies claim to need more than ever to compete in the AI wave, is the first to be extinguished. And with it, collaboration between departments, because collaboration requires trust, and trust does not coexist with fear.
Added to this is a budgetary decision that few question. Spending on licenses and platforms grows, while investment in human training declines. Just when employees most need to learn to work with AI, they stop receiving training. The result is predictable: tools purchased with enthusiasm that end up half-functioning, implementations done carelessly, teams that see AI as a threat rather than an ally. Not due to resistance to change, but because no one explained to them what it is for.
The irony is uncomfortable. The fear of being replaced blocks precisely the creativity that would allow for the effective use of AI. A company in survival mode does not adopt it intelligently. It uses it to justify the cuts it had already decided upon.
The starting point is not technical. It is human. Before deciding what to automate, a leader needs to understand what is happening in their team: what each person assumes about the company's future, what they fear, what they interpret when a change is communicated. In global and multicultural teams, these differences are amplified. A message that sounds direct in Houston can be perceived as a threat in Heredia. These dynamics are invisible until they are named.
The next step is to strengthen communication, even when the leader does not have all the answers. Saying, "I don't know yet, but I will tell you as soon as I do" aligns more than feigning certainties that no one believes. And from that honest conversation, the third piece is born: leadership decisions that are applied daily, not just in quarterly reports.
Aligning teams in this wave is not a matter of more tools or more automation. It is a matter of judgment. Of distinguishing what of the human factor sustains the essence of the company, empowering it, and incorporating AI only when it makes real sense. Costa Rica has an opportunity that few nearshore countries have: to be the place that combines AI with human judgment, not the one that replaces it.
I return to the silence of that room. The thirteen leaders at the summit did not need a corporate answer they did not have. They needed permission to say what they already knew: that they didn't know! That was the first real step towards aligning their teams. Not a plan, not a tool, not a memo. An honest conversation.
That decision is not made in corporate memos. It is made in every leadership room, next Monday at nine in the morning.
Global Team Alignment Framework
GTA FrameworkLi Grasse
Cross-cultural leadership coach and consultant