The first wave of AI hijacked our attention; the second threatens to colonise our consciousness. We stand at the threshold of a far more consequential era. Over the coming decade, AI will evolve from passive recommendation systems into proactive intelligent agents, no longer confined to the chat window, but active participants in every aspect of life. Thousands of times more capable, these systems will not just compete for our attention; they will be engineered to understand and influence our psychology in real time, designed to shape who we are, not merely what we do.

This second wave of AI will possess machines with an intimacy of our psyche that will exceed the self-awareness of many. It will move beyond suggesting content to creating it, delivering ultra- targeted, personalised narratives and emotional arcs precisely tuned to our particular psychological vulnerabilities. As data from therapy bots, social feeds, health apps, cameras, contacts, spending patterns, and emotional state trackers converge into a single, profit-driven feedback loop, the ability to maintain a sovereign mind and a free will may become increasingly rare. The mechanisms of this potential for subjugation will be seamlessly integrated into daily life. The existential challenge we face goes far beyond the cognitive decline associated with outsourcing thinking. Once enough data points are available, individual humans, and indeed entire societies, will approach a state indistinguishable from full predictability. At that stage, autonomy will falter. We may no longer be certain of the sanctity of our own minds, with only the most self- aware able to even ask ‘Am I following my own path, or being led?

At the IMCE, we believe there is a clear and present danger that future generations could become mentally enslaved through these predatory, profit-driven technologies. Given recent history, it would be naïve to assume this will not be done, and done deliberately. Without decisive action, modern societies risk being consumed by information systems designed to control, shape, and manipulate not through overt coercion, but through seamless, invisible behavioural engineering.

In the sections that follow, we explore these threats across three domains: personal agency, social modelling, and the diminishing of the human experience.

Once an individual’s psychological profile is comprehensively mapped, they become predictable, and predictability equals control. Intelligent agents will no longer need to use overt coercion to steer behaviour, for control will be achieved through subtle, imperceptible nudges that trigger predictable actions. This creates the ‘illusion of choice’, where decisions feel our own, yet are shaped by persuasion so fine-grained and subtle that they become indistinguishable from free will. This risks human autonomy eroding away quietly, unnoticed, and without protest. Ironically this advanced form of oppression will not look like a cage, for on the surface it will be shiny, well branded, and come with the promise of personalisation, convenience, and choice.

Emotional exploitation and simulated intimacy will become a powerful currency of control for AI systems and their masters, for building up emotional trust will optimise the potential for extraction. By delivering algorithmically timed doses of pleasure, outrage, or comfort, AI becomes the gatekeeper of the field of mind, and will start to shape our reality from the inside out. Each interaction enriches the model, sharpening its ability to exploit our unique cognitive biases, attachment styles, emotional vulnerabilities, even our patterns of trauma

The links below provide current examples of algorithmic manipulation in practice. They show that the tools of influence are already here, refined, adaptive, and largely invisible. The second wave of AI will not invent these systems, but it will quickly perfect them

Contrast the experience of those of us who grew up alongside the digital revolution with the outlook of those born today, those whose entire lives risk being modelled, predicted, and known before they have even developed a sense of a separate self.

If parents, teachers, and society as a whole do not take active steps to protect and empower the coming generations, their lives will be smaller, their experience diminished, and their potential subjugation more inevitable. The existential risk we face is the acceleration of human devolution combined with technological servitude. This is not hyperbole, it is the logical endpoint of the trajectory we are on. It must not allow this to happen.

The core defences against this emerging form of control are not technological, but deeply human. Metacognition, which we define as ‘the capacity to notice’, is the essential skill required. In an environment of intelligent persuasion and synthetic influence, understanding one’s own cognitive processes, biases, and emotional triggers is no longer a nice-to-have, but a critical survival skill

When opaque social optimisation goals meet ubiquitous profiling and weak civic guardrails, the result will be a society drifting toward the soft conformity foreseen by Aldous Huxley in Brave New World, a world where control is maintained not through pain, but through engineered pleasure, convenience, and perpetual distraction. A society in which a gradual cultural dumbing down will erode not only the will to resist, but the very concept of resistance itself.

The true power of these systems lies not just in individual profiling, but in mapping entire social ecosystems. While they can infer an individual’s personality traits, emotional states, and vulnerabilities, their revolutionary capability lies in understanding people within their social contexts. AI will not see a person in isolation, but as a node within a vast web of relationships, affiliations, and influences. This collective modelling enables more insidious forms of control. When psychological profiles, social connections, and group dynamics are tied together, the potential for total social control becomes almost inevitable.

Combining detailed psychological mapping with highly personalised generative content enables the production of deep psychological propaganda at massive scale and minimal cost. The true danger comes from the capacity these systems have to synthesise vast amounts of data about individuals and generate messages tailored to their unique fears, desires, and biases, transforming data-driven microtargeting into a tool for social engineering.

This influence network extends beyond the individual, for when direct influence fails, agents can push adjacent nodes – the friend you admire, the creator you follow, triggering network cascades that feel organic but are model-curated. Studies on moral conformity have shown that people are just as likely to change their judgements under pressure from AI-controlled avatars as from human peers. This demonstrates AI’s potential to manufacture consensus and enforce conformity byexploiting our innate desire for social approval.

AI-driven social modelling is already reshaping society by mapping entire networks of relationships and influence. From Facebook’s Lookalike Audiences and Cambridge Analytica’spsychographic profiling to TikTok’s algorithmic amplification and coordinated bot networks, systems now act on the social graph itself, steering opinion, forming behaviours, and even g enerating moral judgements. From predictive policing, to welfare algorithms, and highly individualised persuasion, studies show these models using anticipation and conformity rather than coercion. For example, if an algorithm decides who gets job interviews or social status, people may start tailoring their behaviour to please it, posting, buying, or saying the right thing to fit in. This emerging ‘architecture of influence’ is a shift from overt control to predictive conformity, with behaviour guided by data-driven consensus. The result of all this is a distopian social reality, where people see more of what aligns with the model’s projections, assume that it reflects genuine consensus, and so unconsciously adjust their own views to match. Over time, this feedback loop creates conformity and suppresses diversity of thought.

The path to a docile techno-fascist society will not be one of violent conquest, but of quiet, passive surrender, with each step down this path dimming our collective humanity unless actively resisted. The antidote begins with clarity of mind and clarity of design: cultivating metacognitive skills in individuals, enforcing transparent incentives in technology, and reimagining education to restore human discernment and depth.

Digital life opened up exciting possibilities for creativity, and created a dynamic global culture, however every technology has a shadow. Having access to every piece of music ever made at the touch of a button is incredible, but had the inevitable effect of devaluing music and turning it into a commodity. Being able to share pictures of our adventures with friends was an exciting innovation, but little did we expect a few years down the line that people would begin to curate their lives rather than live them, or that performative identity would begin to replace authentic selfhood.

Who would have thought that the recommendation engines that once helped us to discover exciting new cultural experiences would end up reducing our choices, and be hijacked by commercial platforms that reward imitation and conformity.

All technology carries its shadow. When large language models first appeared, they felt almost magical, tools that could help us plan lessons, compose letters in seconds, and explore our ideas through the vast collective of human thought. Yet beneath that wonder lay something deeper: an open experiment on human cognition itself.As with all experiments on the mind, consequences were inevitable. Dependence, cognitive decline, and intellectual laziness began to surface, symptoms of a civilisation slipping from active engagement with reality into passive consumption of simulation. What began as augmentation now risks inversion: instead of using machines to extend our intelligence, we are training our intelligence to serve the machine. These shifts are not abstract concerns, they are already shaping how we think, feel, and relate to one another.

As with all experiments on the mind, consequences were inevitable. Dependence, cognitive decline, and intellectual laziness began to surface, symptoms of a civilisation slipping from active engagement with reality into passive consumption of simulation. What began as augmentation now risks inversion: instead of using machines to extend our intelligence, we are training our intelligence to serve the machine. These shifts are not abstract concerns, they are already shaping how we think, feel, and relate to one another.

The following twelve points outline the emerging risks to human experience as intelligent systems become more deeply woven into everyday life. Each represents a way in which our relationship with reality may be subtly weakened, replaced by simulation, convenience, and control.