Generative artificial intelligence, cognitive plausibility, and institutional responsibility in the era of socio-technical systems
Abstract: The contemporary debate on artificial intelligence can no longer be limited to questioning the abstract nature of algorithmic systems, but must instead analyze what they produce within the social, cognitive, and institutional contexts in which they are used. Drawing on the perspectives of the National Institute of Standards and Technology and UNESCO, the article interprets AI as a socio-technical technology, in which the computational dimension, human practices, cultural expectations, and organizational decisions are deeply intertwined. The central issue is the threshold of cognitive plausibility: generative systems are not conscious and possess no interiority, yet they are capable of producing formally coherent, argumentative, and sometimes empathetic statements that simulate processes human beings recognize as thought. Through reference to Daniel Dennett’s distinction between the physical stance, design stance, and intentional stance, the article highlights the risk of confusing a useful interpretive strategy with ontological proof of the presence of a mind. The problem is not only philosophical, but also cultural and political: the anthropomorphization of technologies can increase trust, delegation, and dependence, altering the way individuals attribute authority, competence, and responsibility to artificial systems. The article concludes by arguing that the future of AI will not depend solely on the evolution of machines, but on societies’ ability to govern their effects through rules, accountability, risk assessment, human oversight, and a sense of limits.
Keywords: #artificialintelligence #generativeAI #cognitiveplausibility #sociotechnicalsystems #anthropomorphization #simulationofthought #Dennett #intentionalstance #NIST #UNESCO #accountability #humanoversight #institutionalresponsibility #AIethics #humanrights #algorithmicgovernance #PaolaLaSalvia #EthicaSocietas #EthicaSocietasJournal #HumanSciences #SocialSciences #EthicaSocietasMagazine #ScientificJournal #ethicasocietasupli
Paola La Salvia: is a former lawyer and senior officer in the Italian Guardia di Finanza, she lectures in economic and legal subjects and is an expert in anti-money laundering and organized crime. A Knight of the Order of Merit of the Italian Republic, she is also the author of several works. Her latest book, I Malacarni, focuses on mafia-related crime. LinkedIn profile.
In the contemporary debate on artificial intelligence, it is no longer sufficient to ask what these systems are in the abstract; it is now necessary to question what they do in the world, how they reorganize practices, expectations, and relationships, and how they transform us as users. This perspective is consistent with the approach adopted by the main international institutions, which describe AI not as a purely technical object, but as a socio-technical technology, deeply intertwined with human behavior, contexts of use, and organizational decisions (NIST, 2023; UNESCO, 2021).
Within this framework lies the role of the National Institute of Standards and Technology (NIST), the agency of the United States Department of Commerce that develops guidelines and standards for the management of technological risks. NIST emphasizes that AI risks emerge from the interaction between technical aspects, organizational factors, and social dimensions, and proposes a risk management model oriented toward systems that are reliable, safe, transparent, explainable, accountable, and non-discriminatory (NIST, 2023). In parallel, UNESCO insists on human rights, dignity, transparency, equity, and human oversight as indispensable conditions for the responsible use of artificial intelligence (UNESCO, 2021).
The most relevant point is the recognition of a threshold that has already been crossed: that of cognitive plausibility. Generative systems are not conscious, nor is it necessary to attribute interiority to them in order to explain their impact; nevertheless, they are sophisticated enough to produce statements that, in form, coherence, and contextual adaptation, simulate processes we recognize as thought. This impression does not arise from a dystopian fantasy in the style of Blade Runner, but from a much more everyday and, precisely for that reason, more insidious effect: models trained to predict the next token are capable of generating responses that appear intentional, argumentative, and at times even empathetic.
Here, a classic question in the human sciences emerges: the distinction between explanation and the attribution of intentionality. Daniel Dennett’s reflection offers a particularly effective interpretive framework, distinguishing between the physical stance, the design stance, and the intentional stance (Dennett, 1987). Adopting the latter means treating a system as if it were an agent endowed with beliefs and desires when this proves useful for predicting its behavior. However, this interpretive effectiveness can generate a cultural short circuit: there is a risk of mistaking a descriptive strategy for ontological proof, ultimately taking the simulation of a mind for a mind itself.
The friction, however, is not merely philosophical: it is profoundly cultural. People already tend to anthropomorphize technologies, attributing warmth, competence, personality, and even conversational norms to them. This phenomenon can facilitate interaction and increase trust, but it also entails the risk of excessive delegation and overestimation of the systems’ capabilities. The problem, therefore, is not only that AI “seems” intelligent, but that this perceptual similarity changes the way we judge, use, and integrate such tools into everyday cognitive processes.
From here emerges the real political issue. The decisive question is not whether machines will become human, but whether institutions will be able to keep pace with their impact. NIST warns that, in the absence of adequate control mechanisms, AI systems can amplify inequalities, produce unfair outcomes, and render automated decision-making processes opaque (NIST, 2023). At the same time, UNESCO reiterates that ethical principles must be translated into concrete policies, effective regulatory instruments, and genuinely operational forms of human oversight (UNESCO, 2021). The debate, therefore, cannot remain confined to the metaphysics of intelligence: it must evolve into a reflection on responsibility, accountability, risk assessment, rights protection, and algorithmic governance.
The decisive point is precisely this: artificial intelligence must be interpreted as a device that redistributes symbolic, cognitive, and institutional power. It does not merely require new technical skills, but demands the development of new interpretive categories. The task is to understand how the authority of discourse, trust in language, the very notion of competence, and the boundary between assistance and substitution, between mediation and dependence, are changing. Maturity, in this context, does not consist in believing or not believing in machines, but in building an ecosystem in which their effectiveness generates neither uncritical adherence nor defensive rejection.
Ultimately, the future of AI will not be decided so much by the intelligence of machines as by societies’ ability to govern their effects with lucidity, measure, and a sense of limits.
BIBLIOGRAPHY
Dennett, D. C. (1987). The Intentional Stance. Cambridge, MA: MIT Press.
National Institute of Standards and Technology. (2023). Artificial Intelligence Risk Management Framework (AI RMF 1.0). U.S. Department of Commerce. https://doi.org/10.6028/NIST.AI.100-1
National Institute of Standards and Technology. (2024). Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile. U.S. Department of Commerce.
UNESCO. (2021). Recommendation on the Ethics of Artificial Intelligence. United Nations Educational, Scientific and Cultural Organization.
LATEST FIVE CONTRIBUTIONS BY THE SAME AUTHOR
AWARENESS BEYOND STRENGTH AS THE FOUNDATION OF WOMEN’s DEFENCE AND FREEDOM
THE PEDAGOGY OF EVIL: WHEN THE ADULT EXAMPLE TEACHES VIOLENCE
MONEY LAUNDERING: THE TRUE BRIDGE BETWEEN THE CRIMINAL ECONOMY AND THE LEGAL ECONOMY
SYRIA: FROM THE GHOSTS OF ASSAD TO A FRAGILE REBIRTH
THE WORLD AT WAR: WHY PEACE HAS BECOME THE TRUE REVOLUTION OF THE 21ST CENTURY
LATEST CONTRIBUTIONS ON AI
THE IMPACT OF ARTIFICIAL INTELLIGENCE ON CORPORATE FINANCE
THE RISK OF RELIGIOUS BIAS IN ARTIFICIAL INTELLIGENCE
THE FUTURE OF FINANCE: THE IMPACT OF ARTIFICIAL INTELLIGENCE ON BUSINESS VALUATION
ARTIFICIAL INTELLIGENCE AND BIAS: LIMITS AND OPPORTUNITIES
GENERATIVE ARTIFICIAL INTELLIGENCE HOW DOES THOUGHT WILL CHANGE?
FIVE LATEST CONTRIBUTIONS
VITAL POVERTY, ALGORITHMS, AND THE PREVENTION OF CHILD SEXUAL ABUSE
BETWEEN SPECTACULARIZATION, EMOTIONAL SUFFERING, AND ARCHETYPES OF PUBLIC SUFFERING
THE SILENT FALL: WHY ANNA DEMOCRITO AND THE OTHERS WERE NOT SEEN IN TIME
ENERGY GEOPOLITICS AND MACROFINANCIAL INSTABILITY
WORK AS A GEOPOLITICAL VARIABLE: MAY 1ST IN THE MULTIPOLAR ORDER
Ethica Societas is a free, non-profit review published by a social cooperative non-profit organization
Copyright Ethica Societas, Human&Social Science Review © 2026 by Ethica Societas UPLI onlus.
ISSN 2785-602X. Licensed under CC BY-NC 4.0