The Application of Artificial Intelligence in Diagnosis and Psychotherapy

RESEARCH ARTICLE The Application of Artificial Intelligence in Diagnosis and Psychotherapy Rahmani Djamel Doctor (PhD) Faculty of social and human sciences; University of Tizi Ouzou Mouloud Mammeri Algeria Email: Senoussaoui Abderrahmen \ \ \ \ \ \ Doctor (PhD) Faculty of social and human sciences ; University of Oran 2 Mohamed Ben Ahmed Azerbaijan Email: ; Doi Serial               Keywords Application, Artificial Intelligence, Diagnosis and Psychotherapy \ Abstract \ As the second decade of the twenty-first century draws to a close, human aspirations are growing to resolve escalating X global challenges—especially those related to mental health. With the limitations of traditional approaches becoming increasingly evident, there is renewed hope that advanced computer technologies, particularly artificial intelligence (AI), may offer innovative solutions. Psychological disorders continue to strain the available resources, making it difficult to meet the needs of a growing number of individuals seeking care. In response, the integration of new, 4 scalable, and more effective methods has become a necessity. Recent advancements in AI applications in public 4 health—particularly in mental health—suggest that AI has the potential to significantly transform the field of psychiatry X for both clinicians and patients. This development raises a critical question: What are the prospects of using artificial X intelligence in diagnosing mental health conditions and providing psychological support to those in need? Citation Rahmani Dj., Senoussaoui A. (2025). The Application of Artificial Intelligence in Diagnosis and Psychotherapy. Science, Education and Innovations in the Context ofModern Problems, 8(6), 290-297; doi:10.56352/sei/8.6.30. ч/ / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / Licensed © 2025 The Author(s). Published by Science, Education and Innovations in the context of modern problems (SEI) by IMCRA - International Meetings and Journals Research Association (Azerbaijan). This is an open access article under the CC BY license . Received: 17.01.2025 Accepted: 30.04.2025 Published: 15.05.2025 (available online)

Contemporary societies are witnessing a significant rise in the prevalence of mental disorders. In contrast, the field of mental health care is facing a clear shortage in the human resources required to provide adequate psychological services. This has led to a substantial gap between the actual need for psychological support and the available capacity to meet that need. Despite ongoing efforts to improve mental health services, the limited number of specialists and the difficulty in accessing treatment—especially in remote or resource-poor areas—remain significant barriers to achieving comprehensive coverage. According to the World Health Organization (2020), mental disorders are among the most pressing public health challenges today. The scarcity of mental health professionals, including psychologists, psychiatrists, clinical psychologists, and psychiatric social workers, constitutes one of the main obstacles to delivering essential mental health care to all.

In this context, artificial intelligence (AI) has emerged as a promising tool that may help bridge this gap by developing more efficient and cost-effective digital diagnostic and therapeutic solutions. Numerous recent studies suggest that AI could bring about a fundamental transformation in psychiatric practice for both clinicians and patients ( Allen , 2020, p. 2). In recent years, AI applications have demonstrated promising potential in supporting psychotherapeutic practices, whether by enhancing diagnostic accuracy through the analysis of behavioral and linguistic patterns or by developing data-driven digital therapeutic models based on big data and machine learning. Additionally, AI-powered chatbots and smart applications have been used to provide immediate and continuous psychological support. The significance of these applications lies in their ability to deliver flexible, customizable, and low-cost services compared to traditional methods.

However, the effectiveness of these tools, the limits of their use, and the ethical concerns surrounding them remain subjects of wide debate and lack sufficient systematic scientific evaluation.

To what extent can AI technologies be employed in psychotherapy to overcome the challenges posed by the shortage of human resources and provide effective and safe psychological support? And what are the ethical considerations associated with this use?

The development of computers experienced a qualitative leap during World War II, particularly with the British secret project Colossus , designed to decode Nazi messages encrypted by the Enigma machine. Mathematician Alan Turing, a key figure in this project, is widely regarded as one of the founding fathers of both computer science and artificial intelligence. In 1948, Turing contributed to the development of the first modern electronic computer at the University of Manchester. Two years later, he published a paper in which he proposed a thought experiment known as The Imitation Game , later referred to as the Turing Test . In this experiment, a third-party evaluator engages in a written dialogue with two hidden participants—one human and one machine—without knowing which is which. If the evaluator cannot reliably distinguish the machine from the human, the machine is considered to have successfully simulated human intelligence. This conceptual model remains a reference point today in evaluating the capabilities of AI systems, particularly in fields such as natural language processing, pattern recognition, and human-computer interaction (Al-Hamwi, 2024).

It was Professor John McCarthy of Stanford University who first coined the term Artificial Intelligence in the early 1950s, defining it as “the science and engineering of making intelligent machines, especially intelligent computer programs” (Nasrallah & Kalanderian, 2019, p. 33).

Previous Studies

Other studies have indicated that commonly used therapeutic approaches—such as Cognitive Behavioral Therapy (CBT), Psychodynamic Therapy (PDT), Psychoanalytic Therapy (PAT), as well as systemic and integrative approaches—can be adapted for digital programming. Acceptance and Commitment Therapy (ACT), MindfulnessBased Therapy (MFT), and Interpersonal Therapy (IPT) have also been incorporated into self-help programs that can be delivered online, either with or without professional guidance (Alimour et al., 2024, p. 2). Additionally, various meta-analyses consistently suggest that Computer-Aided Cognitive Behavioral Therapy (CCBT), delivered via desktop or mobile applications, can be as effective—or even more effective—than traditional face-to-face CBT (Alimour, et al., 2024).

The Meanings of Artificial Intelligence

Machine learning is defined as a set of methods and algorithms capable of automatically discovering hidden patterns in data and then using them to make predictions or decisions regarding future data. In contrast, deep learning is a branch of machine learning based on multi-layered artificial neural networks, which allow computers to learn from past experiences and represent knowledge through a hierarchical sequence of abstract concepts.

Machine learning can be categorized into three main types:

• •        Supervised learning , where the system is provided with pre-labeled data and is used to learn the

mapping relationship between inputs and outputs.

• •        Unsupervised learning , which is applied to unlabeled data to uncover the internal structure or

patterns of the data.

• •        Semi-supervised learning , a hybrid model that combines elements of both previous types, involving a

mixture of labeled and unlabeled data.

Additionally, some researchers categorize AI into two fundamental types:

• •         General (or strong) AI , which is hypothesized to possess analytical capabilities equal to or surpassing

those of humans, including the ability to self-reflect, feel, and be conscious.

• •    Narrow (or weak) AI , which refers to systems designed to perform specific tasks efficiently without true awareness or understanding. All current AI applications are considered examples of narrow AI (Nasrallah & Kalanderian, 2019, p. 34).

Applications of Artificial Intelligence in Psychotherapy Practice :

For most mental disorders, diagnosis and treatment effectiveness in psychiatry still rely entirely on conscious, subjective symptoms and observable clinical signs assessed by qualified human professionals. However, with the advancement of biotechnology and the advent of the era of big data, there are growing notions that AI could serve as a bridge linking psychiatry and psychology with foundational research in various sciences, drug development updates, and global clinical practice advancements. AI methods now have the potential to integrate clinical, psychological, and biological validations and confirm probabilities of accuracy in a digital format (Sun, et al., 2023).

In addition to AI designed to replicate human processes, physicians and researchers have explored the use of animallike smart robots to enhance psychological outcomes such as stress reduction, alleviation of loneliness, emotional regulation, and mood improvement. Companion robots like Paro , a robotic seal, and expressive bear-like robots such as eBear interact with patients and provide benefits akin to animal-assisted therapy. Paro has been used to assist dementia patients experiencing isolation or depressive feelings. AI-powered robots have also been studied for supporting children with autism spectrum disorders (ASD) through educational and therapeutic interventions (Pham, Nabizadeh, & Salih, 2022, p. 251).

Medical and technological researchers are also working on incorporating AI-generated innovations into smart robotic designs for clinical practice. For instance, animal-like smart robots such as Paro —a soft, stuffed polar seal—are increasingly used to support dementia patients. Alongside the furry robot eBear , these are categorized as “companion robots,” functioning as home health aides that interact with users through speech and movement in dynamic dialogues. They aim to provide psychological support to the elderly, isolated individuals, or those suffering from depression by offering interaction and companionship. These robots have shown promise in supporting patients with autism, dementia, and mental disorders by enhancing communication, social skills, and reducing isolation and stress. Despite the initial positive outcomes, these technologies remain experimental and require thorough ethical and regulatory evaluation before widespread clinical adoption (Fiske, Henningsen, & Buyx, 2019).

Other goals of AI in psychological diagnosis and therapy include the use of digital games and smartphone applications. Initially employed for symptom tracking and psychoeducation, digital games have now evolved into full intervention programs. Gamification methods are being used in psychological, social, and cognitive fields to target specific deficits caused by various mental disorders, aiming to restore functionality to impaired domains. These services include behavioral cognitive therapy techniques, behavior modification, social stimulation, attention enhancement, and other psychotherapeutic strategies.

Digital games remain widely appealing due to their accessibility via smartphones. Smartphone applications have also become a prominent AI application, such as the MindLAMP app (Learn, Assess, Manage, Prevent), developed by Torous’ group. It uses smartphones and built-in sensors to understand individuals’ experiences with mental illness and predict recovery through data collection, cognitive and behavioral assessments, GPS tracking, and physical activity monitoring. The BiAffect app employs machine learning algorithms and keystroke dynamics (e.g., typing variability, errors, pauses in messages) to predict manic and depressive episodes in individuals with bipolar disorder (Pham, Nabizadeh, & Salih, 2022).

Recent research has focused on the effectiveness of using smartphones and sensors to monitor mental health, understand illness trajectories, and guide recovery. These smart applications gather diverse data such as questionnaires, geolocation, and physical activity. The research team aims to use machine learning algorithms to predict psychiatric relapses and deliver real-time personalized interventions. Studies have shown that consistent movement patterns are associated with better mental health ( Allen , 2020, p. 4).

Real-time digital data captured from smart devices enhances the potential for promoting both mental and physical health by enabling individuals to engage in continuous self-monitoring linked to their everyday environments. Smartphones and smartwatches, for instance, can collect information on sleep, physical activity, and heart rate— contributing to a wealth of data that helps predict relapses, refine diagnoses, and track psychological states with greater precision. Although these methods have demonstrated effectiveness in psychological diagnosis and care, they have yet to be fully integrated into clinical practice worldwide and remain primarily used in research contexts.

In general, common AI applications in psychological and healthcare settings for patients include, according to Doraiswamy, Blease, and Bodner (2020), the following:

• •        Providing necessary documentation for psychologists and psychiatrists (e.g., continuously updating

medical records).

• •        Evaluating when to refer patients to outpatient versus inpatient psychiatric care.

• •        Analyzing patient information to determine diagnoses.

• •        Analyzing patient data to detect acute homicidal ideation.

• •        Analyzing patient data to detect suicidal ideation.

• •        Aggregating patient information to reach diagnostic conclusions.

• •        Conducting mental status examinations.

• •        Interviewing psychiatric patients in various settings to gather medical history.

Real-time digital data captured from smart devices enhances the potential for promoting both mental and physical health by enabling individuals to engage in continuous self-monitoring linked to their everyday environments. Smartphones and smartwatches, for instance, can collect information on sleep, physical activity, and heart rate— contributing to a wealth of data that helps predict relapses, refine diagnoses, and track psychological states with greater precision. Although these methods have demonstrated effectiveness in psychological diagnosis and care, they have yet to be fully integrated into clinical practice worldwide and remain primarily used in research contexts.

Ethical Issues Related to the Use of Artificial Intelligence in Psychotherapy

• 1.    Regulatory Frameworks : Clear regulatory frameworks must be developed to determine whether, and which, embodied AI applications should undergo conventional health technology assessments and require formal approval from competent authorities. This includes extending provisions that govern the use of such applications outside the direct supervision of healthcare professionals.

• 2.    Professional Guidelines and Training : Professional associations and formal and informal bodies operating in the field of mental health should develop guidelines for the optimal use of AI in this domain. They should also provide recommendations for integrating this topic into medical training curricula to equip future psychiatrists, psychologists, and care providers to engage effectively with embodied AI tools within integrated models of mental and healthcare.

• 3.    Supplement, Not Substitute : AI tools in mental healthcare should be seen as complementary resources—not substitutes—for high-quality specialized care. It is necessary to closely monitor their impact on the availability and use of traditional services.

• 4.    Supervised Use and Risk Management : To uphold care obligations and ensure harm reporting, embodied AI applications should ideally remain under the direct supervision of mental health professionals. Applications used outside formal therapeutic frameworks—such as digital apps and chatbots—must meet reliable standards for risk assessment and ensure appropriate referral pathways to professional care when needed.

• 5.    Transparency and Privacy : Ethical use of such technologies requires transparency, including the development of clear guidelines that ensure respect for patients’ autonomy and privacy—especially regarding obtaining explicit and informed consent for the use of personal data.

• 6.    Algorithmic Bias and Accountability : AI algorithms should be subjected to critical scrutiny, particularly concerning potential biases in data and design. Mental health professionals should be trained to explain the functioning of these algorithms to patients. These technologies should also be open to ongoing scientific discussion and review.

• 7.        Broader Implications : Finally, the expansion of embodied AI in mental health necessitates in-depth analytical

studies to examine its direct and indirect effects on the therapeutic relationship and on human relationships more broadly. Additionally, it is important to assess its influence on individuals’ self-perception, identity, and agency. Longterm consequences must also be considered, such as the rise of reductionist views of health, the increasing objectification of human beings, and the implications of these trends for our understanding of humanity itself.

Conclusion

The primary challenge facing artificial intelligence in mental healthcare is not its efficiency, but rather its integration into daily clinical practice. For these systems to gain widespread adoption, they must be approved by regulatory bodies, integrated with electronic health record (EHR) systems, and standardized sufficiently to ensure the consistent operation of comparable smart products. Moreover, these systems must be incorporated into medical education, funded by public or private institutions, and subject to continuous updates.

While overcoming these challenges will likely take longer than the maturation of the technologies themselves, it is expected that AI will be used in a limited capacity in clinical practice within the next five years, with broader adoption anticipated within the next decade.

It is also probable that AI systems will not replace human clinicians but rather support and enhance their efforts in delivering care. Over time, the role of clinicians may become increasingly centered on tasks that rely on uniquely human skills, such as empathy and persuasion. Conversely, healthcare providers who resist collaboration with AI may find their roles diminished or obsolete over time.