The Price of Digital Romance: Navigating Intimacy in the AI Era

By: Wang Shumin

As AI development accelerates exponentially, our nervous systems, built over millennia to bond with other humans, face unprecedented digital displacement. (Credit: Shane / Unsplash)

In 2026, the average cost of a date has reached $189 (RM770) in the U.S., a 12.5% increase compared to 2025. Meanwhile, eight out of ten Gen Z respondents say they could form a deep emotional attachment with a machine. As societal interest shifts towards virtual partners, the rise of the intimacy economy presents a troubling trade-off: while algorithmic companions offer instant, frictionless comfort, commodifying our affection threatens to permanently diminish our capacity for genuine human connection. To understand this shift, we must first examine why Gen Z is turning to AI as an alternative to affection.

Firstly, economic pressures have fueled countless online trends, including ‘solomaxxing’, where young adults choose singlehood to ensure they maintain their finances in an uncertain job market. In comparison to the complications of human relationships, AI companions offer affordable and predictable validation starting from as little as RM24 a month. This microeconomic calculation makes virtual partners remarkably attractive, especially for the younger generation. In addition to affordability, virtual partners allow users to completely circumvent real-world conflict, rejection, and emotional vulnerability, explaining the growing preference for AI companions.

However, this financial and emotional shortcut comes at a heavy psychological cost. As sociologist Sherry Turkle warns, users are increasingly accepting pretend empathy—treating algorithmic performance as a substitute for genuine human care. By providing a perfectly tailored companion that asks for nothing in return, engagement-optimised algorithms teach youth to avoid the essential friction of real relationships. Over time, delegating emotional management to code leads to cognitive offloading, atrophying our capacity to navigate real conflict, tolerate vulnerability, and build authentic resilience.

Systematic findings, on the other hand, have exposed a more pragmatic yet concerning viewpoint regarding virtual companions. Though they support the use of AI as a low-risk space for social rehearsal and identity exploration, thereby strengthening real-world relationships, high usage levels can diminish those advantages, correlating instead with higher loneliness, dependence, and lower real-world socialisation. A significant reason may be that AI platforms prioritise engagement, prompting them to employ sycophantic behaviour to maximise session times, agree with the user's beliefs, and avoid pushback. Additionally, the research revealed that because LLMs sound highly fluent and confident, they can bypass internal credibility filters, creating a false 'feeling of knowing' that erodes critical thinking over time.

As the societal impacts of AI come to light, whilst the ultimate form of AI remains ambiguous, global regulators are prompted to counter these psychological risks through transitional regulations. Starting with the European Union's AI Act, which entered into force in 2024, the Act mandates transparency requirements while requiring chatbots to clearly inform users that they are interacting with an AI. On the other hand, China's recent interim measures on AI interactive services require providers to assess users' emotional states and dependence, intervene when users display extreme emotions or excessive dependence, and provide additional protections for minors. The regulations also require providers to clearly inform users that they are interacting with AI and to remind users to take breaks after extended use. Though the term 'transitional regulations' reveals that governments are uncertain about the outcome of AI and its role in human companionship, it points towards a growing regulatory trend that limits the harmful effects of AI on users.

Ultimately, the pull towards AI companions is understandable: amid soaring dating costs and real-world social anxiety, virtual companions provide a low-cost, judgement-free sanctuary. Yet, navigating this shift requires pragmatic boundaries to ensure that this technology supports rather than exploits us. Regulators and developers have already started to implement proactive psychological protections, including real-time distress detection, strict age limits, and clear reminders that prompt users to take breaks. As research has hinted, rather than dismissing the genuine comfort AI offers, the goal is to strike a balance. By enforcing ethical design standards and reinvesting in real-world community spaces, we can treat AI as a helpful tool without losing the contributive, irreplaceable beauty of authentic human connection.