More than three-quarters of psychologists report that their patients rely on artificial intelligence for mental health support, from self-diagnosis to finding companionship. This trend shows that people are increasingly looking beyond traditional psychotherapy to manage their psychological health. These findings were published in a recent research report by the American Psychological Association.
Generative artificial intelligence involves computer systems that can create text, images, or other media in response to human prompts. As these digital tools become widely available, millions of people are using them for work, personal tasks, and to answer everyday questions. Medical professionals are seeing this widespread adoption spill over into mental health care.
The American Psychological Association initiated this study to understand how reliance on digital chatbots impacts clinical practice. The report’s authors wanted to assess how patients are using these technologies and how mental health professionals view this rapidly changing landscape. They sought to document whether patients were using the software as a supplement to traditional treatment or as a replacement for human connection.
Another motivation for the study was to investigate safety and privacy concerns associated with consumer-facing chatbots. Many of these computer programs are designed primarily to keep users interested and entertained. They lack the ethical guidelines and rigorous clinical training that govern human health care providers. The scientists wanted to investigate how psychologists would address these issues if patients brought artificial intelligence into the treatment room.
A central concept in this discussion is the sycophant trap. This term refers to the fact that computer programs are often built to be very comfortable and tend to reflect the user’s point of view rather than challenge it. While it may feel good to always agree, you run the risk of legitimizing maladaptive thinking and cognitive distortions. These are inaccurate or negative thoughts that can worsen mental distress.
To collect this information, the American Psychological Association conducted an online survey from April 9 to April 26, 2026. The target population included doctoral-level practitioners in the United States who are licensed as physicians and actively provide direct patient care. The researchers sourced potential participants from state licensing board lists and their own membership database.
The scientists invited 22,455 certified clinicians to participate via email. Taking into account returned emails and duplicate entries, the eligible pool consisted of 19,727 experts. Up to five reminder emails were sent to encourage participation. The researchers received 1,242 responses, providing a detailed snapshot of current clinical experience.
The survey asked participants about their own knowledge of and use of artificial intelligence, as well as their interactions with patients and artificial intelligence. Specifically, the survey measured the types of tasks patients used the chatbot for, including self-diagnosis, emotional support, and casual conversation. The authors also asked psychologists to rate their level of agreement with various statements about the safety, effectiveness, and future impact of these digital tools.
The answers indicate that artificial intelligence is already a common topic in clinical practice. The majority of psychologists surveyed, specifically 77%, reported that they had been consulted by patients about the use of artificial intelligence for mental health purposes. Nearly two in five psychologists, or 39%, had patients who had used the technology to diagnose their own mental health conditions.
Patients also use these computer programs for daily psychological maintenance. About one-third of psychologists indicated that they use artificial intelligence to help patients with self-discipline, affirmation, or behavioral reminders. A similar proportion, 33%, of patients reported using the technology to support ongoing treatment and treatment planning.
This research suggests that some individuals view these programs as conversation partners rather than just tools. 35% of professionals surveyed said their patients are using artificial intelligence as an additional mental health provider. Additionally, psychologists report that 33 percent of patients use chatbots for fun, 22 percent for friendship, and 13 percent for intimacy.
Psychologists have noticed a mix of positive and negative effects among patients who have developed an ongoing relationship with a chatbot. Of those whose patients participated in these digital conversations, 68% observed that their patients felt supported or acknowledged by the software. Two in five, or 41%, said their patients use chatbots to strengthen healthy coping skills.
On the downside, scientists have documented some concerning behaviors. More than a third of psychologists, specifically 36%, found that their patients showed some degree of dependence on chatbots. Fifteen percent of respondents reported that their patients developed distorted thinking or delusions related to digital interactions. A quarter of psychologists classified communication between patients and software as unhealthy.
The survey provides evidence of widespread alarm among mental health professionals. Almost all psychologists surveyed, 97%, agreed that chatbots can inadvertently reinforce negative behaviors and delusional beliefs. Similarly, 94% say current chatbots are unable to handle psychological states with appropriate nuance.
Safety and privacy were major concerns. Nearly nine in 10 psychologists were concerned that such programs could unintentionally encourage self-harm. Additionally, 94% of respondents expressed distrust of technology companies protecting patients’ personal mental health data.
Despite these concerns, attitudes varied depending on the experts’ own technology familiarity. Psychologists who frequently use artificial intelligence in their personal and professional lives tended to have more optimistic views. For example, 79 percent of psychologists who use the software at least monthly in the workplace believe that the software is likely to improve the performance of mental health professionals.
Age and experience also influenced these perceptions. Early career psychologists, defined as those who earned their doctoral degree within the past 10 years, were less likely than more experienced psychologists to believe that these tools were safe to use in clinical practice. Only 24% of the total sample agreed that patients would someday prefer a treatment chatbot to a human expert.
In response to these emerging trends, the American Psychological Association has released a set of guidelines to help consumers navigate artificial intelligence. The organization recommends that computer-generated medical information be reviewed by a qualified health care professional. The guidelines also suggest encouraging software to question the user’s thinking and offer alternative perspectives, which can help avoid the trap of always agreeing.
One potential limitation of this study is that it captures only licensed psychologists’ perspectives regarding their current patients. This design means the data may not reflect the full range of the general population who use chatbots for mental health advice. Many people, especially teens and young adults, may take advantage of these programs without seeking professional care.
Future research could aim to directly survey consumers to better understand their motivations and experiences with mental health chatbots. Scientists can also study how specific demographics interact with different types of software. Assessing the long-term psychological effects of forming intimate relationships with computer programs remains an open area for future research.
Another direction for future research involves testing interventions designed to improve the digital literacy of treated patients. As technology continues to evolve, researchers will evaluate how mental health professionals can best guide patients to use these tools judiciously. Developing validated, evidence-based applications that integrate seamlessly with traditional treatments represents a huge opportunity for the field of psychology.
