It’s the day before your biggest exam of the semester! You spent hours cramming in the library, dropped hundreds of additional dollars on note-publishing services, and drained your Starbucks Card on much-needed fuel. Still, you can’t escape the test anxiety that inevitably washes over you.
However, maybe there’s something you could do to stay above that crashing wave. Sure, you could call your mom or vent to your roommate, but why even consider connecting with a real human when there’s an accessible, technological powerhouse at your fingertips: an AI-powered chatbot.
This may sound completely absurd to you, but what if I told you that Honorlock, a leader in online exam proctoring software, actively promotes this anxiety-reducing method on their blog. Yes, the very software that actively blocks AI during your exam is suggesting that you use AI as a confidant and emotional anchor.
But beyond this irony, a bigger question can be raised:
Are young people actually turning to AI instead of each other, and if so, should they?
Well, for starters, AI for mental health isn’t merely a listless coping recommendation. Instead, it’s a tool that’s being actively used by young people. In fact, according to research organization RAND’s 2026 survey, “19.2% of young people ages 12 to 21 said they had used AI chatbots” for emotional support, which is “up from 13.1% in a similar RAND survey conducted a year earlier and is similar to the 19.8% who reported receiving counseling from a mental health professional.” If this growth continues, we could very well observe an increase in the number of people who utilize AI chatbots as emotional support systems in lieu of actual trained therapists. On one hand, when compared to formal therapy, which some individuals may find to be expensive or time-consuming, AI provides the advantages of instant access and often free use… as long as your conversation doesn’t get cut off because you ran out of your daily premium credits.. On the other hand, a number of key points should be taken into consideration before readily adopting AI as a new personal, digitized shrink.
1. How could AI venting impact us?
Out of the AI therapy users included in the RAND 2026 survey, “92% said the advice was somewhat or very helpful.” But the funny thing about life, and particularly the human condition, is that our individual thoughts and perceptions may not always reflect the reality at hand or the deeper issues at play. Sure, a large majority of the survey participants might have found AI advice to be helpful in the moment, but the quality of said advice is truly what dictates its usefulness in the long term.
According to the Stanford Report, the quality of AI advice is indeed in question. Researchers found that AI models were highly sycophantic in advice contexts, meaning the systems were “overly agreeable,” endorsing users’ thoughts significantly more than real people. Furthermore, it was noted that these sycophantic responses were perceived as more trustworthy than non-sycophantic ones. Thus, it seems plausible to conclude that AI therapy may function as an echo chamber of sorts, in which AI chatbots seek to please the user by providing advice that aligns with what the user already wants to hear. This leaves little-to-no room for substantial growth, because, unlike with real human therapy and connection, AI advice-seekers may not be challenged to reflect or consider new perspectives. They aren’t being probed to look deeper to identify the root of their feelings, and the AI doesn’t have feelings to care about understanding the user further.
So, if we live in a world where young people using AI as therapists is the norm, a world where they receive only the advice they want to hear and deem trustworthy, we may be losing out on the chance to grow from honest feedback. Instead, we’re just attempting to fix complex issues with athe figurative Band-Aid of words that feel good.
2. How could AI venting impact our relationships?
In addition to AI therapy’s hindrance to true growth, overreliance on AI venting could take prime conversational real estate away from topics that would normally bring us closer to others. For instance, venting and advice sharing are key social tools that we use daily to facilitate relationships. According to 2009 research by psychologist Bernard Rimé, when people experience emotion, they discuss it with another person 88%–96% of the time, with 60% of those conversations happening on the same day as the emotional incident. Gossiping, venting, recounting, yapping—you name it—isn’t for just pure fun or convenience. It’s woven into who we are and how our connections with others grow. If we begin to consistently utilize AI chatbots to placate our emotions, how will we use our feelings to support our bonding with others?
This sentiment connects well to the fact that AI use has been correlated with higher levels of loneliness. According to MIT’s Media Lab, “Those who spent more time typing or speaking with ChatGPT each day tended to report higher levels of emotional dependence on, and problematic use of, the chatbot, as well as heightened levels of loneliness.” AI robs an individual of a two-way conversation in exchange for accessibility and an echo chamber. We shouldn’t rob ourselves of using our emotions as a tool for social/emotional growth simply because AI provides immediate gratification.
So, where does that leave us?
At its core, AI is a tool that many young people, especially students, have been encouraged to learn about and utilize for their benefit. We use it to draft emails, build career plans, and generate practice tests. We learn that it’s a system of probabilities and tokens, and that “AI won’t replace our jobs, but people who know how to use it will.” These are helpful concepts and good initiatives to take amidst this technological revolution.
Although many individuals consider AI to be an innovative and useful tool, we must tread lightly and avoid relying on AI chatbots to solve our feelings and worries. There are clear limitations that call into question AI’s ability to provide quality emotional support, as AI often produces sycophantic outputs that inspire baseless trust. Additionally, over-relying on AI could distance us from real-world support systems and temporarily satiate feelings that need deeper probing or human support.
Finally, there lies a concerning dichotomy, with sources like Honorlock recommending AI as a coping tool, while MIT produces research on how those same chatbots are linked to loneliness.
We’ve barely scratched the surface on social media’s impact on mental health, let alone the long-term consequences of becoming dependent on algorithms for mental support. If we begin to depend on AI for emotional support, we may forget the real reason why human connection matters so much. We don’t share things with others to simply be heard and have our thoughts regurgitated back at us with a couple more em dashes. We share with others to be understood, to look at a person, a friend, a family member—and know that they are capable of not only hearing us, but feeling what we feel too.