The convergence of Artificial Intelligence (AI) and mental healthcare signifies a major transformation in healthcare. The growing awareness of mental health as a crucial aspect of overall well-being has highlighted the global mental health crisis. The World Health Organization (WHO) indicates that mental health disorders contribute substantially to the global disease burden, with depression leading the cause of disability worldwide. The increased prevalence of mental health disorders has placed significant strain on healthcare systems, exposing the inadequacies of traditional models that primarily rely on in-person consultations and therapies. These traditional models struggle to meet the rising demand for accessible, affordable, and scalable mental health services. AI offers potential solutions due to its capabilities in handling large datasets and identifying complex patterns. In mental healthcare, where understanding complex human behaviors and emotions is crucial, AI can revolutionize care by providing insights and solutions unattainable through conventional methods. AI offers advanced detection approaches, personalized therapies, and virtual therapeutic platforms, potentially increasing access to care, reducing stigma, and improving treatment outcomes. However, this integration also presents ethical considerations, regulatory challenges, and the need for ongoing research and development. The synergy between human expertise and AI capabilities holds the potential to usher in a new era of mental healthcare.

This narrative review examines the use of AI in mental healthcare, focusing on papers published in peer-reviewed journals, conference proceedings, and reputable online databases between January 2019 and December 2023. The review traces AI's integration into mental healthcare from the mid-20th century, highlighting milestones such as the emergence of cognitive modeling and the development of ELIZA, an early chatbot simulating a Rogerian psychotherapist. The evolution continued with expert systems, computerized CBT programs, and ultimately, the diverse applications of AI in the 21st century, including early detection of mental health issues, personalized treatment plans, virtual therapists, advancements in teletherapy, and continuous monitoring. The review also explores the history of AI in mental health, from early cognitive modeling to modern AI-driven interventions, demonstrating the growing recognition of technology's potential to support mental well-being.

This paper uses a narrative review approach to comprehensively investigate the use of AI in mental healthcare. The inclusion criteria for paper selection included papers published in peer-reviewed journals, conference proceedings, or reputable online databases that specifically focus on the application of AI in mental healthcare, including review papers offering a comprehensive overview, analysis, or integration of existing literature. Exclusion criteria included papers not meeting these criteria, duplicates, non-English publications, or those unrelated to the review topic. The screening process involved three stages: title screening, abstract screening, and full-text eligibility assessment. The search strategy aimed to identify relevant papers published on "Artificial Intelligence in Mental Healthcare" between January 2019 and December 2023 across four databases (PubMed, IEEE Xplore, PsycINFO, and Google Scholar). A total of 211 papers were initially identified, with 92 eligible studies included in the final review after excluding duplicates and non-English language papers.

The review found that AI is playing an increasingly significant role in various aspects of mental healthcare. **Diagnosis:** AI aids in the early detection and prediction of mental health disorders by analyzing speech, text, facial expressions, and electronic health records (EHRs). Examples include the use of NLP to analyze text and speech for indicators of depression or anxiety, and facial expression analysis to identify subtle cues. AI also facilitates predictive modeling, considering multiple factors to assess an individual's risk of developing a mental health condition. Tools like Woebot, Cogito, and Affectiva were highlighted for their applications in this area. **Treatment:** AI is transforming mental health treatment through personalized interventions and virtual therapists/chatbots. AI algorithms analyze individual characteristics, including genetic predispositions and behavioral patterns, to customize treatment plans. AI can also predict a patient's response to different treatment approaches. Virtual therapists and chatbots offer increased accessibility and scalability of mental health support, exemplified by platforms like Woebot, Wysa, Talkspace, and BetterHelp. **Therapy Delivery:** AI enhances teletherapy by analyzing patient emotions in real-time during video sessions, providing therapists with valuable insights to adjust their approach. AI also assists therapists by analyzing large datasets, suggesting interventions, and automating administrative tasks. Examples include Kintsugi and Cerebral. **Monitoring and Follow-up:** AI-enabled wearable devices continuously monitor physiological and behavioral markers related to mental health, allowing for early detection of relapses or deterioration. AI-driven outcome assessments provide objective measurements of treatment progress, guiding data-driven decisions and enhancing treatment efficiency. Examples include Oura Ring and Mindstrong Health.

The integration of AI in mental healthcare offers promising solutions to address the global mental health crisis. AI-powered tools provide personalized interventions, improve early detection, and offer accessible virtual therapy platforms. However, ethical considerations must be carefully addressed. Privacy and data security are paramount, requiring robust measures to protect sensitive patient information. Bias mitigation is crucial to ensure fair and equitable access to care across diverse populations. It is also important to maintain the human element in therapy, using AI as a tool to augment rather than replace the therapeutic relationship. The need for clear regulatory frameworks and transparent validation of AI models is essential for responsible implementation.

AI holds significant potential for enhancing mental healthcare by improving accessibility, personalization, and efficiency of care. However, responsible and ethical implementation requires addressing challenges related to privacy, bias, and the preservation of the human element in therapy. Future directions include establishing clear regulatory frameworks, ensuring the validation and transparency of AI models, and investing in ongoing research and development to fully realize AI's potential in transforming mental health care. Collaboration among stakeholders, including clinicians, researchers, policymakers, and technology developers, is crucial for successful and ethical integration.

While the review highlights the considerable potential of AI in mental healthcare, some limitations exist. The sensitive nature of mental health data necessitates stringent privacy protections. Algorithm bias poses a risk of inequitable access and treatment. AI lacks the human empathy and understanding crucial in therapeutic relationships. Integration with existing healthcare systems and navigating regulatory complexities pose further challenges. Future research should focus on addressing these limitations to ensure the responsible and effective implementation of AI in mental healthcare.