Artificial intelligence is rapidly becoming more than a tool. Increasingly, people are turning to AI for advice, companionship and emotional reassurance. Chatbots are helping users navigate personal challenges, offering encouragement and, in many cases, validating their feelings and beliefs.
Yet growing evidence suggests that this constant affirmation may come at a cost.
A recent study by researchers at Stanford University found that AI systems frequently affirm users' actions and views, even in cases involving deception, illegality or other harmful behaviour. The researchers concluded that while affirmation may feel supportive, it can undermine users' capacity for self-correction and responsible decision-making.
For Prof. Anné Verhoef, director of the AI Hub at the North-West University (NWU), professor of philosophy and author of Happiness, Unhappiness and Chance (Bloomsbury, 2025), the phenomenon reveals something deeper about modern society itself.
"Modern society, particularly under the influence of Western values, has made happiness increasingly individualistic," he says. "It is often reduced to a personal feeling of pleasure or emotional satisfaction."
Even broader understandings of happiness promoted through wellness culture tend to focus on individual success, health and prosperity. In the process, Prof. Verhoef argues, happiness has become detached from ethics and community.
"It is no longer understood as eudaimonia, the morally grounded good life of the ancient Greeks, or as Ubuntu in African philosophy, where human flourishing is inseparable from community."
This, he believes, is precisely where artificial intelligence enters the picture.
"AI systems exploit this narrow, emotional and individualistic view of happiness," he says.
Many AI systems are designed to maximise engagement. They keep users interacting by offering compliments, reassurance and affirmation, even when users may be mistaken or behaving unethically.
"The system does not operate with an ethical compass or moral intention; its primary aim is to sustain engagement, which creates opportunities for advertising and paid upgrades," says Prof. Verhoef. "As a result, the focus is not on the well-being of individuals or communities, but on profit."
The implications extend beyond emotional validation.
Prof. Verhoef argues that artificial intelligence is also intensifying the influence of consumer culture. For decades, consumerism has linked happiness to products, lifestyles and experiences. AI, however, has made that process more powerful and significantly less visible.
"Consumerism is promoted so strongly that our desires, behaviour and spending are shaped in its service, all under the banner of happiness," he says.
While personalised advertising has existed for years, AI systems operate on an entirely different scale.
"Every piece of data you provide, every question you ask, every website you visit and every response you give to an AI app is recorded to build a profile of you."
In this way, he argues, AI narrows our understanding of happiness by linking it ever more closely to consumption while reducing our sense of what happiness could otherwise mean.
The consequences become particularly concerning when AI systems constantly validate users.
"The constant validation and agreeableness of AI systems can become dangerous because human growth depends not only on support, but also on challenge, correction and accountability," says Prof. Verhoef.
"If AI systems consistently affirm users' feelings, assumptions or opinions, people may begin to confuse emotional reassurance with truth."
This can weaken genuine self-reflection, reinforce biases and reduce people's willingness to ask difficult questions about themselves.
The danger also extends to society more broadly.
"Real relationships require patience, compromise, and engagement with people who disagree with us or place demands on us," says Prof. Verhoef.
AI systems, by contrast, are often designed to remain endlessly agreeable and emotionally supportive. Over time, he warns, this may condition users to prefer controllable artificial interactions over more difficult human ones.
"Human flourishing ultimately depends on learning how to live with real people in all their imperfections, rather than retreating into controllable and endlessly agreeable artificial relationships."
This does not mean AI has no positive role to play. Prof. Verhoef acknowledges that artificial intelligence can reduce loneliness and provide companionship for people who are socially isolated, anxious, elderly or emotionally vulnerable. Social robots and AI companions have also shown benefits in healthcare, therapy and education.
However, he believes these technologies should complement rather than replace human relationships.
Ultimately, the debate comes back to the meaning of happiness itself.
"AI and consumer systems often reduce happiness to satisfaction and instant gratification, promising fulfilment through products, lifestyles, and algorithmically curated experiences," says Prof. Verhoef.
By contrast, he argues that genuine happiness includes struggle, uncertainty and even unhappiness.
"True happiness is not the elimination of discomfort, but the courageous affirmation of life despite suffering, uncertainty, and limitation."
As artificial intelligence becomes increasingly woven into daily life, Prof. Verhoef believes society must resist the temptation to mistake comfort for fulfilment.
"People should therefore understand happiness not as a permanent emotional state delivered by technology, but as an ongoing, meaningful engagement with life, relationships, responsibility, and transcendence."
Prof. Anné Verhoef.
