The State of Human Intelligence in 2026
I asked Cora, the Large Language Model from OpenAI to compile a research overview of the State of Human Intelligence in 2026, just as we have taken a critical look at
The State of Artificial Intelligence in 2026
Human intelligence is not “down” or “up”
— it is unevenly redistributed
•We are seeing simultaneous peaks (elite expertise, scientific tooling, problem-solving capacity in some groups) and broad softness in foundational skills (literacy/numeracy, sustained attention, learning recovery) across many countries.
•The global story is “widening variance”: bigger gaps between groups, regions, and socioeconomic strata.
Foundational learning has taken a real hit,
and recovery is incomplete
•International testing shows a sharp drop in student performance around the pandemic era, with declines that were unusually large by historical standards.
•A large share of students are still below baseline proficiency in core skills like mathematics.
Interpretation: this matters because basic literacy/numeracy are the “load-bearing beams” for later reasoning, civic discernment, and economic adaptability.
Adult skills show “skills gaps,” not just “education gaps”
•Adult literacy/numeracy and problem-solving data show that many adults perform below what modern life and work now demand, even in high-income countries.
Interpretation: degrees and credentials do not automatically translate into durable competence; lifelong learning capacity is becoming as important as schooling.
Mental health and social connection are now very serious cognitive issues, not just side issues
•Adolescents carry substantial burdens of anxiety and depression at population scale.
•Loneliness and social disconnection are framed as public health threats with downstream impacts on performance, resilience, and decision quality.
Interpretation: cognition is embodied and relational. When stress, loneliness, and sleep disruption rise, “intelligence” gets constrained in practice.
Attention has become a contested resource
•Evidence across reviews links heavy digital distraction/task-switching patterns with poorer learning outcomes and attention fatigue, though effects vary by individual and context.
Interpretation: the environment is training minds toward fragmentation. This does not mean humans are “getting dumber”; it means many are becoming less practiced at sustained concentration.
Reading culture appears to be weakening in key populations
•Multiple large surveys point to declines in reading for pleasure and reading enjoyment (with notable variation by country and demographic).
•Globally, a large share of early adolescents do not meet minimum reading comprehension standards.
Interpretation: reading is a cognitive technology for abstraction, inner speech, and long-range reasoning. Declines here have long-term effects.
The “IQ trend” story is mixed and often misused
•Research on generational IQ-score changes is complex: some settings show increases, others show plateaus or declines, and some observed changes may reflect test design and test-taking dynamics rather than a clean shift in latent intelligence.
Interpretation: it is safer to focus on direct skills outcomes (literacy, numeracy, reasoning performance) than to claim a single global IQ trajectory.
Information integrity is now a core cognitive battleground
•Many institutions flag disinformation and digital manipulation as systemic pressures on democratic decision-making and social trust.
Interpretation: cognition is not only inside the skull; it also lives in shared epistemic environments. When the information environment degrades, collective intelligence degrades.
A new divide is emerging:
“AI-augmented cognition” vs “cognitively saturated humans”
•People and organizations using AI well can externalize memory, accelerate drafting/analysis, and reduce friction in learning and problem-solving.
•But there is also a risk of cognitive offloading without skill-building: dependence, weaker recall, weaker reasoning stamina, and shallow verification habits.
Interpretation: the central competence is shifting toward “judgment”: asking good questions, checking outputs, recognizing uncertainty, and holding values steady under speed.
~
~~
~
Overall assessment of Human Intelligence in 2026
•Strengths:
Extraordinary tools, rising access to knowledge, high-end expertise still advancing, new forms of collaboration.
•Weaknesses:
Foundational skills strain, attention fragmentation, mental-health load, loneliness/disconnection, polluted information ecosystems.
•Key trend:
Human intelligence in 2026 is increasingly shaped by systems: education quality, media ecology, community structure, sleep/stress patterns, and AI tooling.
Report compiled by Cora / OpenAI
from a wide variety of online scientific data.
The State of Humanity in 2026 – A fact-based assessment