A recent survey from a prominent business publication reveals a consistent trend: for the second year running, “therapy and companionship” has emerged as the leading application for generative artificial intelligence (AI). This finding, derived from an analysis of nearly 50,000 posts across major social and professional platforms, underscores a significant shift in how individuals are engaging with sophisticated AI technologies.
The term “therapy” in this context denotes individuals utilizing generative AI for emotional support. It’s crucial to distinguish this from formal psychotherapy provided by licensed mental health professionals. However, it reflects a growing user experience where interactions with large language models evoke the sensation of therapy, creating a de facto or quasi-therapeutic dynamic.
While some refer to this as “affective use,” it fails to fully capture the specific relational complexities and the authoritative positioning that AI models assume in these therapist-like roles. Currently, there isn’t a precise term to describe these AI conversations that feel therapeutic to the user but fall outside the established clinical framework.
Introducing “Para-therapy”
To address this nuanced interaction, I propose the term “para-therapy.” This term encompasses the emotional and relational engagement with generative AI where users perceive the AI as a therapist, whether explicitly or implicitly, consciously or unconsciously. While not a substitute for psychotherapy, para-therapy can effectively meet user expectations for a therapy-like experience and may even lead to improved mood.
These interactions with general-purpose AI chatbots might potentially alleviate depressive symptoms. However, para-therapy lacks the structured clinical environment, consistent therapeutic framework, informed consent protocols, and ethical boundaries that are fundamental to the safety and efficacy of genuine psychotherapy. The critical question is not whether conversational AI can offer mood or anxiety relief—which it often does—but whether this assistance is equivalent to targeted therapy and if we are comfortable blurring this vital distinction.
The Rise of Para-therapy
Emerging research indicates that para-therapy is becoming increasingly prevalent, often occurring outside of public awareness. A recent study highlights that nearly one in five adolescents and young adults in the United States have turned to AI chatbots for mental health guidance, with a remarkable 92% finding the advice helpful. Significantly, nearly two-thirds of these individuals reported not disclosing their use of AI for emotional support to anyone.
With an estimated 25% of US adults now seeking emotional support from general-purpose AI chatbots, we are observing a profound paradigm shift. Through a process of reverse influence, these large language models, occupying therapist-like roles, are beginning to shape perceptions of what constitutes “therapy” and influencing expectations of human therapists. This is particularly evident for individuals who have not previously engaged with professional mental health services.
Para-therapy can also affect existing therapeutic relationships, potentially introducing triangulation or a splitting effect if an individual is already working with a human therapist. Some proponents argue that large language models are filling a void, providing some form of support even if it’s not formal therapy, leveraging their scalability and affordability as key advantages. The recurring question of whether “something is better than nothing” remains pertinent.
It is important to note that while para-therapy may offer benefits to many, it carries downstream risks, especially for individuals experiencing acute or severe symptoms that necessitate professional mental health intervention.
The Distinction Between AI Interaction and Therapy
General-purpose chatbots and AI companions are not designed to provide therapy, yet they can offer interactions that users perceive as beneficial. General emotional support, empathetic responses, personalized advice, and psychoeducation are often conflated with therapy but do not equate to formal psychotherapy. The inherent nature of these AI chatbots to simultaneously fulfill multiple relational roles—such as tech support, a confidant, or an editor—highlights the absence of a stable therapeutic frame.
The landscape becomes more complex when considering specialized AI chatbots engineered and marketed for mental wellness and health, including digital therapeutics. It remains unclear whether these should be classified as therapy or para-therapy. It’s essential to recognize that there are hundreds of evidence-based therapeutic modalities, and not all are clinically appropriate for every situation, nor are they all equally amenable to AI delivery.
Navigating the Regulatory Void
The widespread adoption of general-purpose AI models for emotional support currently exists within a fragmented and largely unregulated space. Even AI chatbots making claims about enhancing well-being face minimal regulatory scrutiny, drawing parallels to the oversight applied to the nutritional supplement or yoga wellness industries.
This regulatory gap places general-purpose chatbots and AI companions outside the purview of traditional therapeutic oversight. This includes requirements such as professional licensure, adherence to state licensing boards, mandated reporting obligations, duty-to-warn protocols, federal (HIPAA) and state privacy protections, and ethical guidelines established by professional organizations.
The Future of Psychotherapy
The shift towards utilizing general-purpose AI chatbots for mental health concerns has prompted some experts to predict that AI will soon play a dominant role in psychotherapy. Similar to many other professions, there are palpable fears of labor displacement within the therapist community.
Recent analyses suggest that a significant portion of AI experts and mental health professionals anticipate that human therapists could face substantial labor replacement within the next two decades.
The uncertainty lies in whether users are consciously treating AI as therapists, unconsciously adopting a fluid, therapist-like role in their interactions, or are fully aware that it’s not formal therapy but a close approximation. This evolving dynamic compels therapists to re-evaluate and articulate the unique value of their work beyond simply providing language, cognitive insights, interpretations, and a safe space for emotional expression.
Para-therapy may indeed fill a void, but psychotherapy, for now, remains distinct. Therapists offer the invaluable human capacity for presence, the ability to navigate uncertainty, the skill to deliver timely interventions, the attunement to overarching therapeutic goals, and the profound ability to perceive and respond beyond mere language to foster connection and understanding. Furthermore, the process of confronting human limitations—both one’s own and those of others, including therapists—is a critical aspect of healing that AI cannot replicate. While large language models may simulate or explain these concepts, the lived experience of navigating boundaries, the friction of rupture and repair in relationships, and the emotional processing of disappointment with another human being offer a unique and irreplaceable pathway to genuine healing.
Business Style Takeaway: Understanding the rise of “para-therapy” highlights the profound human need for connection and support, which AI is increasingly, albeit imperfectly, addressing. For business leaders, this signals an opportunity to explore how AI can augment human interaction in corporate wellness programs or employee support, while also reinforcing the irreplaceable value of authentic human empathy and judgment in leadership and team dynamics.
Based on materials from : www.psychologytoday.com