AI and Technology in Grief Support: Clinical Implications and Ethical Considerations

Application of AI-Driven VR in Individual Therapy Settings

AI-driven VR technology might be used in individual therapy as an aid to grief counseling, where the client is given an opportunity to “see” the avatar of their deceased loved one and potentially interact with them in the immersive virtual environment (Pacaol et al., 2024). Although unconventional, this approach might offer a novel therapeutic intervention that would allow the client to tap into their connection with the deceased more readily, accessing memories of them and their relationship. As discussed earlier, VR technology’s application to grief relates closely to the continuing bonds theory (Klass et al., 2014), highlighting the potential adaptive strategies to connect with the deceased person to process one’s grief. Although grieving individuals can engage in various practices to maintain sense of a connection with the deceased (e.g., visiting their favorite places, listening to music that reminds them of the deceased, browsing through photos, writing unsent letters, or creating meaningful rituals to stay connected with the deceased), AI-driven VR technology may offer a more immediate potentially profound means of eliciting emotional responses related to the deceased (Pizzoli et al., 2023). AI-driven VR-assisted therapy, with its immersive nature, may be particularly beneficial for grieving individuals who struggle to connect with their emotions or who exhibit significant emotional avoidance in acknowledging the reality of loss.

Given its immersive and interactive nature, VR has also been investigated as a therapeutic intervention for complicated grief with potential applications as an adjunct to or alternative modality for traditional therapy. In a longitudinal study, clients who participated in VR-based therapy sessions—including psychoeducation, emotional processing, and narrative reconstruction—demonstrated significant reductions in complicated grief symptoms and depression (Quero et al., 2019). While both VR-based and traditional in-person therapy showed lasting improvements at the 12-month follow-up, the VR condition demonstrated larger effect sizes, suggesting that VR may enhance therapeutic outcomes by providing a structured yet emotionally engaging environment for processing grief (Quero et al., 2019).

Furthermore, VR could also be used in cases where grieving individuals have a hard time processing the circumstances of the death, which is not uncommon in situations where the death is sudden, traumatic, or has other complicating factors (Fisher et al., 2020). In these cases, and similar to the PTSD treatment using graded-exposure VR technology (Heo & Park, 2022), VR may be used as a graded exposure technique to help the surviving person reexperience aspects of the loss as a way to gradually integrate the reality of how the death happened. Even though this approach holds promise, it is also important to note that future research should examine how the VR-assisted exposure technique may compare to the usual talk therapy approach, where the client works on gradually accepting the reality and circumstances of the death. It is possible that VR-assisted grief counseling may be too immediate and, thus, too intense or otherwise maladaptive to some grieving individuals. Although research in this area remains extremely limited, one study presented preliminary efficacy data in comparing a VR-based CBT intervention, Emma’s World, to standard talk-based CBT in a sample of participants diagnosed with stress-related disorders, including pathological grief, PTSD, and adjustment disorders (Baños et al., 2011). The findings demonstrated that the Emma’s World VR intervention was as effective as traditional CBT and concluded that VR-based approaches may serve as a viable alternative for clients who do not prefer conventional treatment modalities (Baños et al., 2011).

VR may enhance widely utilized grief counseling interventions, such as the empty chair technique (Paivio & Greenberg, 1995; Sharbanee & Greenberg, 2023), letter writing (Larsen, 2022) or serve as another way to address any “unfinished business” and facilitate closure for grieving individuals (Paivio & Greenberg, 1995). By providing a realistic visual representation of the deceased and enabling a meaningful “encounter,” VR can transform traditional grief counseling interventions (e.g., Cheng, 2025; Kim et al., 2020, 2024). Instead of imagining the deceased in an empty chair or mentally visualizing them while writing a letter, clients can engage in these exercises by interacting with an AI-generated avatar of the deceased through VR. Such technology may extend letter writing into a verbal narrative directed at the deceased or, with advanced AI, facilitate a back-and-forth simulated conversation. Although these interactions could be a powerful tool for connecting with the deceased and may evoke stronger emotional reactions, their therapeutic value in the grieving process remains to be explored. Given the lack of empirical research on the utility of immersive VR in grief therapy, further research is needed to determine its impact on bereavement adjustment and to inform grief counseling practice, following established guidelines for VR-based therapeutic applications (Birckhead et al., 2019).

Application of AI-Driven VR in Group Therapy Settings

AI-driven VR could also be integrated into group therapy settings. Similar to individual therapy, grief group participants could use VR to engage in the aforementioned therapeutic interventions. Although techniques such as the empty chair (Paivio & Greenberg, 1995; Sharbanee & Greenberg, 2023) and letter writing or narrative conversation (Larsen, 2022) could be conducted in a group setting, completing these exercises in a more private space on their own may be more feasible. In a world where all group members have access to VR tools, they could engage in these activities privately while still benefitting from processing their experiences with other grieving individuals and a trained group facilitator. In addition, discussing these experiences within the group may further support individuals in engaging with continuing bonds, as members reflect on their connections with the deceased, share fond memories, and explore how they sustain these relationships in their daily lives—an approach that is already integrated into traditional group therapy settings (e.g., Larsen et al., 2023). In fact, accessing group support may be particularly valuable as individuals explore using novel tools like VR, providing a safe and supportive environment for shared experiences.

In a controlled study, participants in an eight-week VR-based support group reported greater reductions in depressive symptoms and increased social support compared to those accessing the grief website that utilized readings from a publicly available website (www.HelloGrief.org; Knowles et al., 2017). VR’s immersive nature allows grieving individuals to interact in real time while maintaining anonymity (Knowles et al., 2017). Given that traditional grief support groups are typically accessed by a fraction of individuals experiencing grief (e.g., Steiner, 2006), the use of such technology would be particularly helpful for grieving individuals to access the support they may need, particularly for those experiencing stigmatized and disenfranchised grief or social isolation (e.g., Doka, 1999; Robinson & Pond, 2019). Considering that grieving individuals appreciate online grief support communities for the emotional support they receive from others with shared experiences (Hartig & Viola, 2016; Robinson & Pond, 2019), VR grief support groups could further enhance connections among grieving individuals, serving as a valuable resource to help them feel more connected and facilitate their grieving process.

Emerging research continues to explore the feasibility of AI-integrated immersive VR in grief therapy, particularly for structured interventions targeting grieving individuals. A study by Zhang et al. (2024) evaluated the acceptability of a generative AI-integrated immersive VR program designed to incorporate meaning-centered grief therapy for bereaved parents. This program seeks to facilitate both individual- and group-based meaning-centered grief interventions, beginning with bereaved parents and potentially expanding to other relationships. As part of its development, researchers conducted an initial acceptability survey with 27 bereaved parents, finding that the majority of the participants were open to using VR for bereavement support and interested in engaging with the prototype once available (Zhang et al., 2024). These preliminary findings highlight the potential of VR for grief support in group settings, although further research is needed to elucidate whether such interventions are deemed safe, feasible, and efficacious (Zhang et al., 2024).

Application of AI-Driven Self-Help Tools in Grief Support

With the increasing reliance on digital tools for mental health support, self-help mobile apps, and online grief-related resources have become valuable tools for individuals coping with grief (Krysinska & Andriessen, 2013; Vanderwerker & Prigerson, 2004). For example, mental health professionals recently developed My Grief, a mobile application designed to support grieving individuals through psychoeducation, self-monitoring of grieving symptoms, and CBT-based self-guided exercises (Eklund et al., 2021). A recent study demonstrated that bereaved parents who utilized this app reported high satisfaction, stating it helped them better understand their grief reactions, gain knowledge of prolonged grief, monitor their grieving process, work through their grief, and feel more comfortable seeking support (Eklund et al., 2024). In addition, the Grief Works: Self Care and Love app (PSYT, 2024), developed in collaboration with grief psychotherapist Dr. Julia Samuel, also provides psychoeducation, practical exercises, and interactive tools for grieving individuals. Furthermore, researchers have recently developed an AI screening system for prolonged grief disorder, highlighting its emerging potential in clinical settings; this system assists in the initial assessment of bereaved individuals by evaluating their symptoms and needs and providing personalized results (She et al., 2022).

Currently, AI technology companies catering to grieving individuals primarily focus on developing self-help tools. In fact, companies within the digital afterlife industry (Figueroa-Torres, 2024; Öhman & Floridi, 2018) have prioritized individual-based, self-directed grief support tools over therapist-driven interventions that integrate AI technology. With the boom of AI-driven grief tools, it is not surprising that more and more grieving individuals would have access to this alternative support. Beyond advanced technologies such as AI-driven immersive VR, other emerging tools—including AI-generated chatbots designed to replicate the demeanor and communication style of the deceased—are becoming readily accessible to the general public (Figueroa-Torres, 2024). These chatbots, often referred to as ghostbots or griefbots, offer a realistic, simulated platform where grieving individuals can engage in conversations with their deceased loved ones, ask them questions, seek advice, share about their day, or find ways to address any unsaid or unfinished matters (Xygkou et al., 2023).

Recent research has begun to explore the role of AI chatbots in grief support, examining their potential benefits and associated risks. A qualitative study employing semi-structured interviews explored the motivations, perceived benefits, and potential risks associated with chatbot use among 10 individuals at various phases of bereavement (Xygkou et al., 2023). Participants reported using chatbots in different ways, including simulating interactions with the deceased, talking with a supportive “friend,” and, in some cases, forming romantic attachments to the chatbot. The results indicated chatbots were helpful in “filling the void” left by the deceased and alleviating feelings of loneliness. Some participants described using chatbots to address unfinished business, seek closure, and process feelings of regret and guilt, emotions commonly associated with grief. Further, participants reported that chatbots functioned as nonjudgmental and unbiased listeners, while others highlighted their role in facilitating a continuing bond with their deceased loved ones. Notably, a subset of participants shared that the chatbot became their virtual “romantic partner,” as they felt their social skills precluded them from seeking human relationships at the time. Furthermore, one particularly interesting finding was that most participants explicitly stated that chatbots should not replace real human connections but rather serve as a supplemental tool alongside traditional grief support and therapy. Even tlthough participants recognized that chatbots are not real humans, they nonetheless acknowledged their potential as valuable aids in navigating the grief journey (Xygkou et al., 2023).

Although AI chatbots may provide a sense of comfort by facilitating connections with deceased loved ones, which may be adaptive in some ways (Pearcy, 2023; Xygkou et al., 2023), their impact on the grieving process remains unclear. Concerns persist regarding the safety and ethical implications of these individual-geared grief support technologies, as well as the extent to which they are regulated and whether they are intended for short-term or long-term use. Furthermore, suppose these tools are used independently without the supervision of a licensed mental health provider. In that case, they may exacerbate distress and heighten potential risks, especially for individuals experiencing intense or complicated grief.

Potential AI Applications in Personalized Grief Support

Since virtual environments are not constrained by time and space, AI-driven technology can provide grieving individuals with more timely and responsive support. Psychologists could potentially develop a comprehensive, theory-driven, and evidence-based application integrating AI-powered grief counseling bots, targeted psychoeducation, tailored coping strategies, and personalized resources that extend beyond merely simulating communication with the deceased. Specifically, such application could leverage advanced AI algorithms to continuously monitor and assess individual progress by analyzing user responses, linguistic patterns, cognitive functioning, emotional expressions, physiological indicators, and stress biomarkers, all of which serve as critical indices of grief responses (e.g., Kowalski & Bondmass, 2008; Latham & Prigerson, 2004; Masferrer et al., 2017). By incorporating multimodal data sources, the system could provide customized feedback and interventions while strictly adhering to ethical guidelines.

Developing an AI-driven grief support application would facilitate the delivery of personalized, readily accessible grief interventions, offering more responsive and accessible assistance to those navigating the complexities of grief. Moreover, implementing simultaneous data analysis to identify high-risk individuals with potential suicide risk or substance use—based on their behavioral patterns, communication styles, cognitive functioning, and mental states (e.g., Latham & Prigerson, 2004; Masferrer et al., 2017)—could be particularly helpful in enhancing support for grieving individuals. This approach would enable mental health professionals to initiate early interventions and deliver timely, targeted support that meets the specific needs of grieving individuals, helping to prevent grief from progressing into complicated grief or prolonged grief disorder (Eklund et al., 2021; She et al., 2022).

AI and Culturally Responsive Grief Research

As AI-driven grief support tools continue to evolve, psychologists committed to social justice and culturally responsive mental health interventions are well-positioned to advocate for their development in ways that ensure inclusivity, representation of diverse populations, and accessibility (APA, 2017b). Despite the growing integration of AI into mental health care, discourse on its impact on grief counseling currently remains limited. Psychologists can provide critical expertise to ensure these tools offer safe, person-centered, and equitable support.

Generative AI, a type of AI that uses deep learning to generate new content by learning underlying patterns from large datasets, has faced scrutiny over ethical concerns, mainly due to its potential to produce biased or harmful content (e.g., Chen et al., 2023; Scott et al., 2023). These challenges have driven the development of human-centered AI (HCAI), which prioritizes human control to enhance human abilities, prioritize addressing societal needs, and ethical AI development (Geyer et al., 2022). HCAI emphasizes transparency, fairness, privacy, and inclusion, seeking to mitigate the biases AI systems often inherit from historical and systemic discrimination (Bragazzi et al., 2023; Chen et al., 2023; Scott et al., 2023).

Although the primary purpose of developing technology is to transform and improve individuals’ lives, it is also crucial to ensure that the direction of technological advancement aligns with meeting societal needs and thus adheres to HCAI. Psychologists can leverage AI to advance culturally responsive research and address biases embedded in generative AI models. Since AI is primarily trained on internet data, which often lacks diverse representation, a human-centered approach is essential to prevent the reinforcement of systemic inequities across racial, ethnic, gender, and other identity-based lines (Bragazzi et al., 2023; Chen et al., 2023; Obermeyer et al., 2019). The necessity for HCAI stems from instances where AI-driven decisions have exacerbated societal inequities based on race, ethnicity, gender, age, sexual orientation, disability, class, and religion (e.g., Bragazzi et al., 2023; Chen et al., 2023; Field, 2024; Scott et al., 2023). For example, a widely used U.S. hospital algorithm demonstrated racial bias, requiring Black patients to report more severe symptoms to receive the same level of care as White patients (Obermeyer et al., 2019). Given AI’s reliance on biased datasets (Chen et al., 2023), psychologists are crucial in guiding AI development to reflect diversity, equity, and inclusion (APA, 2017b). Counseling psychologists can help identify and rectify discriminatory AI tools, ultimately mitigating health disparities among marginalized populations (Akbar & Parker, 2021; Chen et al., 2023).

Moreover, as datasets on the internet are primarily in English, addressing language disparities and developing inclusive AI tools that reflect broader linguistic diversity is imperative (Murray, 2024). Utilizing smaller, targeted datasets and sustainable language models can enhance the feasibility of cross-cultural and cross-national grief research while expanding linguistic diversity in AI to improve accessibility in grief support across cultures (Chen et al., 2024; Ngomo, 2024). This approach would enable grief researchers to capture diverse grief experiences, including culturally specific expressions, facilitating the development of culturally sensitive and responsive interventions.

Finally, integrating AI into grief research and clinical practice can expand access to culturally tailored grief support and advance health equity by facilitating the development of culturally responsive and inclusive interventions (Jora et al., 2022; Murray, 2024). Moreover, AI-driven VR technologies have the potential to enhance existing cross-cultural comprehensive grief counseling training by providing immersive, experiential learning that refines clinical skills, improves therapeutic outcomes, and better prepares trainees—including graduate students and mental health professionals—to deliver adaptable and inclusive support across diverse cultural contexts (e.g., Jankauskaite et al., 2024; O’Brien et al., in press). By ensuring that AI development adheres to HCAI principles, psychologists can play a pivotal role in creating ethical, inclusive, and effective tools that address the diverse needs of grieving individuals.

AI and Continuing Bonds in Grief Support: Opportunities and Challenges

As AI and technology introduce new modalities for maintaining connections with the deceased, redefining the parameters of continuing bonds within an AI and technological context becomes essential. Although AI-based interactions offer novel opportunities for grief support, they also raise critical concerns regarding the authenticity and psychological implications of engaging with virtual representations of the deceased (Klass et al., 2014). The phenomenon of sensory and quasi-sensory experiences after loss further complicates the distinction between reality and simulation, necessitating a clearer understanding of how AI influences individuals’ perceptions of their relationships with the deceased (e.g., Kamp et al., 2020). Given that AI recreations do not restore the deceased but instead construct a new interactive entity, it is imperative to establish psychological and ethical guidelines to ensure that these tools facilitate adaptive coping rather than fostering an illusion of relational continuity. Moreover, as grieving individuals engage with virtual representations of the deceased, psychologists must consider the implications of these quasi-attachments—bonds formed through interactions with AI agents that mimic but cannot replicate authentic human bonds—and the potential risks of substituting AI-based interactions for the emotional integration of loss.

Above all, one of the crucial elements of using AI and technology is that the immersive experience may not fade over time, suggesting that using AI and technology to maintain bonds with the deceased would yield different results for grieving individuals’ psychological outcomes. In this regard, it would be important to consider that interacting or maintaining continuing bonds with the deceased through AI and technology is not merely reconstructing the reality one desires to relive. Moreover, the use of AI and technology can be considered as advanced tools that may facilitate grieving individuals who struggle with their grief and find maintaining bonds with the deceased helpful in their grieving process. Many grieving individuals who have utilized AI and technology to “meet” the deceased have found it helpful in working through unresolved grief (e.g., Kim et al., 2020, 2024). This highlights the importance of understanding how AI and technology influence perceptions and understandings of maintaining continuing bonds with the deceased and whether they foster adaptive coping mechanisms.

Posthumous Consent and the Autonomy of the Deceased

A critical yet underexplored ethical issue is posthumous consent–the absence of explicit permission from the deceased to be “revived” through AI. Although AI holds promise as a supportive tool, it is essential to consider the privacy and autonomy of the deceased. In accordance with the APA Principle of Respect for People’s Rights and Dignity, psychologists have a responsibility to protect individuals’ privacy, confidentiality, and self-determination, including in posthumous contexts (APA, 2017a). Thus, similar to existing advanced consent mechanisms for organ donation or social media memorialization accounts, ethical AI grief interventions should implement opt-in consent mechanisms that enable individuals to participate voluntarily in decision-making processes regarding posthumous digital recreation.

Privacy, Confidentiality, and Data Protection in AI

Researchers have recently highlighted the need for legal provisions to regulate AI resurrection services to ensure that AI-generated avatars are protected under personal data laws (Cheng, 2025). In alignment with the APA Principle of Respect for People’s Rights and Dignity (APA, 2017a) it is essential to implement legally recognized consent frameworks, such as explicit opt-in mechanisms during one’s lifetime, ethical oversight committees to review AI-generated representations, or authorization protocols granting decision-making power to designated digital stewards or other authorized representatives. These measures are critical not only for ensuring a balance between the rights of the deceased and grieving individuals but also for safeguarding privacy and confidentiality, particularly as the reuse of personal data in AI systems raises concerns about unauthorized access and potential breaches (e.g., Cheng, 2025). Establishing clear regulations would help mitigate ethical concerns while providing guidelines for the responsible use of AI in grief support.

In accordance with the APA Principle of Beneficence and Nonmaleficence (APA, 2017a), ethical guidelines should be developed to ensure that data utilization for research and practice aligns with the principles and values of the field of psychology. Given the obligation to avoid and minimize foreseeable harm, as well as the APA Principle of Fidelity and Responsibility (APA, 2017a), psychologists can play a crucial role in leading discussions and engaging in research and advocacy on data usage, regulation, protection, and HIPAA compliance, particularly regarding AI models using data from grieving individuals. This involvement ensures that the data of grieving individuals participating in research and practice can be protected and utilized for their intended purpose. Further, it is crucial to address how confidentiality boundaries in research and clinical settings could be managed. For example, suppose individuals utilizing AI-based clinical services disclose information that breaches confidentiality, such as indicating harm to themselves. In that case, immediate intervention is necessary to delineate how such data should be handled to prevent harm to individuals and provide prompt, appropriate support.

Another crucial ethical consideration is the absence of clear guidance on data usage. Since deceased individuals cannot provide consent, and the parameters governing the use of their data for AI development remain unclear, it is imperative to establish robust protections at every stage, from data collection to preservation and AI model training. This includes incorporating human-centered AI principles when collaborating with interdisciplinary researchers and advocating for legislation that ensures privacy and confidentiality, especially as data usage may extend beyond the purview of existing regulations such as HIPAA (McCoy et al., 2023).

Algorithmic Bias and Discrimination in AI

The development of AI technology based on human-created data raises concerns about potential discrimination that would be mirrored in society. Specifically, AI users may experience biases and discrimination related to age, race, ethnicity, gender identity, sexual orientation, national origin, immigration status, language proficiency, religious/spiritual identities, disability, socioeconomic status, neurodiversity, and/or other factors (Chen et al., 2024; Obermeyer et al., 2019; Scott et al., 2023). These underscore the ethical responsibility of psychologists to critically assess AI-driven interventions and advocate for fairness in AI applications. In accordance with the APA Multicultural Guidelines (APA, 2017b), psychologists must actively address systemic inequities and ensure that AI-based mental health technologies do not disproportionately harm marginalized communities (e.g., Chen et al., 2024; Scott et al., 2023). Given that AI models are often trained on datasets reflecting societal biases (e.g., Scott et al., 2023), psychologists play a critical role in advocating for algorithmic transparency, equity-centered model evaluation, and bias mitigation strategies in AI development (Cruz, 2024; Robinson et al., 2024).

To uphold these ethical and multicultural commitments, psychologists must advocate for preventing discrimination in AI-based research, practice, and interventions. Specifically, a critical examination of ethical considerations in AI development, such as algorithmic bias, algorithmic transparency, employing deep learning, informed consent, and privacy protection, can provide essential guidance for researchers and practitioners integrating AI technology in psychological contexts (e.g., Chen et al., 2024). Such research is crucial for safeguarding the well-being and privacy of individuals seeking support through AI-based interventions, which is equally important as protecting those served in the therapy space.

Defining and Establishing Ethical and Clinical Boundaries for AI in Grief Counseling

The establishment of clear guidelines on the appropriate use of AI-based grief interventions is imperative for ensuring the ethical and effective integration of AI in grief support while mitigating potential risks and optimizing therapeutic outcomes (e.g., Robinson et al., 2024). Specifically, AI should be conceptualized as a supplemental tool designed to enhance, rather than replace, human-led interventions (Pizzoli et al., 2023). AI-driven technology, when employed within stringent ethical and clinical guidelines, holds the potential to augment screening, assessment, diagnosis, and initial therapeutic or supportive interventions. Such interventions would facilitate accessible and timely support while bridging systemic gaps, ultimately fostering a more ethically grounded and equitable approach to grief care.

To implement these guidelines effectively, it is essential to have clear decision-making frameworks that define the scope and limitations of AI-driven services. These frameworks should enable AI to accurately assess the level of care required—determining whether self-guided support is sufficient or if professional psychological intervention is needed—while also ensuring that psychologists can monitor and intervene as necessary to provide timely and appropriate support (e.g., Pizzoli et al., 2023). Clear assessment and diagnostic frameworks are essential to ensure that individuals with complex grief presentations, such as complicated grief or prolonged grief disorder, receive the specialized care they need. This would enable the detection of severe grief-related distress, potentially allowing for early or timely interventions while ensuring that mental health professionals can facilitate appropriate interventions (Pizzoli et al., 2023). Furthermore, guidelines should incorporate safeguards against excessive dependence on AI, emphasizing its role as a complementary tool rather than a substitute for grief counseling.