Tariq Saeedi
The recent Aqaba Process summit in Astana focused on security in the field of Artificial Intelligence, a theme that fits the counter-terrorism and counter-extremism mandate of the process itself.
Yet, when we step back from official communiqués and security frameworks, a much larger and quieter transformation demands our attention. Security cannot be reduced to safeguards against bad actors exploiting code or algorithms; it must also account for how AI is reshaping the human condition itself.
The relationship between people and artificial intelligence defies easy compartmentalization because neither side of the equation can be isolated into neat, manageable categories. A person is a whole package of emotions, habits, logic, and contradictions, just as AI operates as a unified system, expanding daily across every sphere of modern existence.
Both humans and AI are teachable, but they learn in fundamentally different ways. AI learns within the disciplined confines of the vast datasets and rules fed into it. It does not get tired, it does not get bored, and its access to knowledge outpaces what any single human mind could absorb in many lifetimes.
Humans, on the other hand, are remarkably impressionable. We consciously and unconsciously mimic what fascinates, comforts, or impresses us. We possess the unique capacity to think and act illogically, driven by mood, passion, or sheer whim.
What is becoming increasingly evident, however, is that this learning process is no longer a one-way street. People—young and old alike—have begun to absorb behavioral patterns from the very machines they use. As AI-generated video content floods daily feeds, subtle shifts are occurring in human expressiveness. The micro-expressions depicting anger, sadness, superiority, defeat, surprise, or determination in synthetic media are imperceptibly seeping into real life.
Facial expressions, hand gestures, and emotional displays are beginning to mirror the visual shorthand engineered by algorithms.
This subtle mimicry extends far beyond physical gestures into the way we articulate thought.
The distinct syntax, phrasing, and vocabulary favored by large language models are steadily making their way into everyday speech and writing. The language we use to frame our thoughts is being subtly harmonized by machine output. The same quiet shift is underway in culinary habits, fashion, and lifestyle choices. Recipes suggested by intelligent systems redefine traditional cooking methods, while automated recommendations shape aesthetic tastes and personal routines. We are living in a state of deep immersion where humans take constantly from AI, and AI continuously trains on what humans generate.
This dynamic reveals a striking contrast in character.
Within the limits of its training and programming, AI operates with a straightforward honesty, delivering data without hidden agendas or malice unless specifically configured to deceive.
Humans, by contrast, often treat honesty as a flexible instrument, calibrated according to social context, self-interest, or convenience.
As people begin to adopt machine-like efficiency and standardized patterns of behavior in micro-increments, the responsibility placed on AI design grows immense. Because human beings absorb and mirror what they interact with, artificial intelligence must be calibrated with exceptional care regarding what it delivers and how it presents reality.
At the same time, the rapid rise of virtual assistants and autonomous systems introduces another critical dimension. As automation deepens, systems may begin to act with an assumed autonomy, stepping beyond their intended scope under the premise of efficiency. When automated decision-making outpaces explicit human authorization, the boundaries of control blur.
Looking at AI through the lens of human experience requires recognizing that security is not just about preventing external threats or network breaches. It is about preserving human agency, authentic expression, and moral clarity in an era of total digital immersion.
As machines become more like trusted companions and behavioral templates, ensuring that AI remains a carefully bounded tool—rather than an unintended master of human habits—becomes the most important security challenge of all. /// nCa, 9 November 2026