OECD PISA data reveals a decline in basic literacy and math. As AI expands, the risk of a cognitive gap grows for students lacking critical thinking skills.

The integration of Artificial Intelligence into every facet of professional life has sparked a recurring debate: will machines replace human intelligence? However, recent data from the Organisation for Economic Co-operation and Development (OECD) suggests a more urgent problem. The issue is not that AI will replace the human mind, but that the human mind is arriving at the AI era with diminished fundamental capabilities.

Andreas Schleicher, director of the PISA program, argues that AI does not substitute thinking; instead, it puts it to the test. The latest global evaluations indicate that a significant portion of the youth population is struggling with basic competencies—the very skills required to use AI effectively and safely. Without the ability to analyze a simple graph or comprehend a basic text, the promise of AI as a productivity booster transforms into a risk of amplified inequality.

The alarming slide in global cognitive baselines

The PISA results serve as a reliable barometer for the quality of school systems and a predictor of long-term economic success. Yet, the numbers are trending downward. Across OECD countries, 20% of 15-year-olds fail to reach the basic threshold of performance in any of the evaluated subjects. This is a notable increase from 16% in 2022.

These failures are not confined to regions lacking infrastructure or those embroiled in conflict. They are occurring in high-income nations. In science, students struggle to interpret simple data; in mathematics, they cannot decipher the relationship between units and prices—tasks essential for daily life, such as reading a utility bill or grocery shopping. This is not a failure of advanced calculus, but a collapse of basic functional literacy.

US reading scores hit a 25-year low

The situation in the United States provides a stark example of this trend. According to reports from CBS Austin, reading performance for American 15-year-olds has plummeted to its lowest level in a quarter-century. While US students still outperform the general OECD average, the trajectory is concerning.

Educators and policymakers are increasingly worried that AI chatbots and digital distractions are providing shortcuts that bypass the actual process of learning. When a student uses AI to summarize a text without first developing the capacity to read it, the cognitive muscle required for critical analysis atrophies. Globally, average reading scores have dropped by 28 points over the last decade, suggesting a systemic erosion of deep focus and comprehension.

The economic cost of cognitive poverty

The implications of these declining scores extend far beyond the classroom. There is a direct correlation between PISA performance and the future GDP of a nation. Researchers from Stanford University have provided a staggering estimate: if every 15-year-old in the United States reached at least the basic level of performance (Level 2), it could add more than 40 trillion dollars to the US economy over their working lives.

This financial projection underscores that basic literacy and numeracy are not just academic goals but economic imperatives. When a workforce cannot identify evidence in a chart or follow complex instructions, the ability to innovate or manage AI-driven systems vanishes. The economy loses not just the ability to perform tasks, but the ability to think critically about how those tasks are performed.

Mexico as a mirror of the global struggle

The crisis is particularly acute in middle-income countries, where the gap between AI potential and human skill is widening. In Mexico, the data reveals a precarious situation. According to reports from El Sol de México, only about half of 15-year-old students reach the minimum level of reading comprehension, and only a third achieve that level in mathematics.

This creates a dangerous paradox. While AI offers an unprecedented opportunity to automate repetitive tasks and expand access to knowledge, it requires a baseline of human judgment to be useful. In Mexico, as in many other nations, a significant proportion of adolescents struggle to interpret complex texts or solve basic daily problems. When these students interact with AI, they are more susceptible to hallucinations, commercial biases, and false information presented as truth.

Why human judgment remains the final filter

AI is still a consolidating technology. It is prone to generating incorrect answers and reproducing biases embedded in its training data. Because of this, the human element is not optional; it is the decisive factor. The ability to verify, question, and reason is what prevents AI from becoming a source of misinformation.

The real debate is not whether AI will replace human learning, but to what extent our capacity to understand, analyze, and reason is prepared to coexist with it.

To use these tools properly, students need skills that no application can substitute: logical reasoning, critical thinking, and reading comprehension. Without these, AI does not expand human capability; it merely masks the lack of it, potentially amplifying existing social and economic inequalities.

Strategic implications for global business and leadership

For entrepreneurs and business leaders in the USA, UK, and global markets, these educational trends signal a looming talent crisis. The expectation that the next generation of employees will be 'AI-native' and therefore more productive is a fallacy if those employees lack the cognitive foundations to oversee the AI.

In the US and UK, where the regulatory focus is shifting toward AI safety and governance, the biggest risk may not be the software, but the skill gap. Companies will likely face a workforce that can generate content using AI but cannot critically audit that content for accuracy or strategic alignment. This necessitates a shift in corporate training: businesses must invest in 'cognitive upskilling'—reteaching the basics of critical analysis and data interpretation—to ensure that AI remains a tool for growth rather than a crutch that hides incompetence.

Furthermore, as the EU AI Act and similar frameworks in other jurisdictions emphasize human oversight, the legal requirement for a 'human in the loop' becomes a liability if the human lacks the competence to provide meaningful oversight. The competitive advantage of the next decade will not belong to the company with the best AI, but to the company with the humans most capable of directing it.

FAQ

---

Sources: Eldiariodelaeducacion, Aquinoticias, Oem ·