The modern corporate job market is undergoing a silent but profound shift. Office workers now have an invisible advantage that their superiors often fail to perceive or objectively assess.
The rapid penetration of Artificial Intelligence (AI) into everyday work is not only changing the way results are produced, but is also fundamentally redefining the concept of professional competence, creating an unprecedented illusion about the real skills of human resources.
The rise of the “virtual” experience
The findings of a recent survey by the Use.AI platform on a sample of almost 10,000 employees (9,684 people) in Europe, the US, Canada and Latin America reveal the extent of the phenomenon. 52% of respondents openly admit that using AI tools makes them appear much more experienced and knowledgeable than they actually are.
Technology acts as an invisible catalyst:
- Empowerment: 64% of employees have used AI to complete tasks they could not have done on their own.
- Taking on responsibilities: 43% have taken on tasks they were not trained to do.
- Fear of exposure: 35% say they would struggle to complete their current tasks without AI, while one in four worry that their employer is counting on a skill that doesn’t actually exist.
- This discrepancy between delivered results and actual knowledge is often left in the dark. 39% of employees have submitted work produced with TN without declaring it, while 30% have taken credit for results produced almost entirely by the technology. With most organizations lacking a clear disclosure framework, career progression is directly affected: 19% of respondents say that AI assistance was instrumental in securing a promotion.
The Managers’ Dilemma
The obvious response—mandating disclosure of every interaction with AI—is deemed unsustainable.
Instead of a surveillance regime, a functional threshold of substantive assistance is proposed:
- When disclosure is required: When AI structurally shapes content (e.g., producing a critical part of an analysis, a proposal, a presentation, or code).
- The goal: Providing the manager with sufficient context to assess both the quality of the output and the human contribution to it.
- The principle of responsibility: The employee submitting the work remains solely responsible for understanding, verifying, and defending it.
- The skillful use of AI is undoubtedly a new, valuable professional skill. However, the final deliverable ceases to be a reliable indicator of individual competence. The crucial question that emerges is whether technology functions as a power multiplier or as a crutch that weakens independent thinking.
The risk of structural dependence becomes visible when the employee:
- Produces an answer through AI but is unable to identify possible errors or omissions.
- Unable to explain the reasoning path and structure of the result.
- Unable to form a solid judgment when the system provides incomplete or incorrect data.
- As Herasimov emphasizes, if gains in speed and volume of work are accompanied by the gradual atrophy of critical thinking, then businesses are called upon to manage a dangerous compromise.
The New Architecture of Assessment in the Job Market
In an environment where producing an impressive document requires minimal effort, employers are looking for new models of performance assessment. As the final product loses its evidentiary value, attention is shifting to deeper human abilities.
The most fundamental skill in the new era is concentrated in problem definition:
- The ability to formulate the right question
- Questioning implicit assumptions
- Documenting the superiority of one strategic choice over another
At the same time, managers are asked to consider whether the employee can diagnose the weaknesses of the answers produced and make decisions under conditions of uncertainty. The distinction between what someone simply “produces” with the help of technology and what they truly “understand” becomes the central pillar for promotion and compensation decisions.
Scaling Autonomy and the Future of Work
While organizations struggle to adapt their assessment systems, technology is advancing at an exponential rate. According to data from METR (via Ventureburn), the time it takes to double the autonomy of AI systems has shrunk from eight months to just 4.7 months.
From early 2025 to early 2026, we saw:
- 1,400% increase in autonomous capabilities of AI systems.
- Explosive rise in AI tool downloads from 15,000 to 11.8 million (a 780x increase).
- 35x increase in MCP (Model Context Protocol) tools, reaching 177,000.
- The proliferation of standards like MCP allows AI systems to connect directly to applications and data sources, moving from simply describing tasks to autonomous execution (agentic AI).
This transition exacerbates a fundamental dilemma. The more a production process is performed autonomously, without inherent human guidance at every step, the less the final result reveals about the true capabilities of the person signing it. The future of work will not belong to those who simply use technology to appear competent, but to those who retain the critical capacity to question it, direct it, and ultimately surpass it.
