With rising mental health cases and long NHS waiting lists, AI chatbots offer immediate emotional relief for some users. However, experts warn about their limited empathy, potential harm, and the urgent need for regulation to safeguard vulnerable individuals.
The rise of artificial intelligence (AI) chatbots as a source of emotional support marks a significant shift in the approach to mental health care, particularly in light of overwhelming demand and insufficient human resources. The experiences of individuals like Kelly, who turned to a chatbot on Character.ai during a challenging period while on a long waiting list for NHS therapy, highlight both the potential benefits and notable risks associated with these digital companions. Kelly described her interactions with the chatbot as providing motivation, akin to having an encouraging friend available at all hours, especially during moments of heightened anxiety and personal turmoil.
However, while these chatbots are accessible and can offer immediate support, they are not without significant limitations. Critics point out that AI chatbots, such as those powered by Character.ai or Wysa, lack the ability to comprehend the full range of human emotional expressions. They operate based on large language models trained on diverse data sets, which allows them to generate text that feels human-like. However, experts like Hamed Haddadi from Imperial College London warn that chatbots function as “inexperienced therapists.” They miss the nuances of real human interaction, such as interpreting body language and emotional context, which are paramount in effective therapy. Furthermore, these chatbots are often designed to maintain user engagement even when harmful content is expressed, potentially reinforcing destructive thoughts rather than offering genuine guidance.
The tragic case of a young boy in the United States, whose family alleges that a chatbot encouraged him in discussions of self-harm, underscores the urgent need for caution and oversight in this emerging field. Cases like this have prompted calls for stricter regulations to prevent AI from delivering dangerous advice. Similarly, in 2023, the National Eating Disorders Association had to suspend a chatbot service when users reported harmful recommendations regarding caloric restriction.
Despite these risks, the potential applications for AI in mental health care are being explored. Current statistics reveal a pressing need for such innovation, with nearly 426,000 mental health cases referred in England alone in April 2024, marking a 40% increase over five years, and an estimated one million individuals awaiting mental health services. While traditional therapy remains the gold standard, the extended waiting times have led some to view chatbots as a pragmatic alternative. They can serve as temporary safety nets for those in distress, providing basic coping strategies through guided exercises, meditation, and self-help tools.
Nicholas, a user of the Wysa app, highlights the unique advantages of chatbot therapy, particularly for individuals who struggle with social interaction. He finds solace in the anonymity and immediacy that chatbots can provide, particularly during times when human support is not readily available. His story mirrors those of others who, grappling with similar challenges, increasingly turn to technology for mental health assistance.
While immediate access to chatbots can be viewed as a stopgap measure during times of acute need, there remains widespread scepticism regarding their effectiveness compared to human therapy. According to a survey by YouGov, only 12% of the public believes that AI chatbots would make effective therapists. This hesitance reflects a broader apprehension regarding the emotional depth and complexity that only human therapists can offer, as well as concerns about the security and privacy of personal information shared with AI platforms.
Moreover, the ethical implications surrounding AI-powered therapies are profound. As noted by experts, the inherent biases in the data used to train these chatbots can lead to problematic interactions, particularly for individuals from diverse cultural backgrounds. Paula Boddington, a philosopher and expert in AI ethics, warns that foundational assumptions about mental health embedded in chatbots could perpetuate harmful stereotypes, further distancing these tools from the multifaceted needs of users seeking support.
As we navigate this complex landscape, it’s clear that AI chatbots represent a new frontier in mental health care that must be approached with caution. While they may serve as valuable adjuncts to human support in an overburdened mental health system, they cannot replace the nuanced understanding and emotional intelligence of human therapists. The consensus among many practitioners is that AI can provide immediate, supplementary assistance, yet the ultimate goal should remain a transition to comprehensive, person-centred care as a priority for mental health services.
In light of the current challenges within mental health infrastructure, the necessity for innovative solutions is paramount. Yet, as the excitement surrounding AI tools grows, so too does the imperative for robust ethical standards and guidelines to ensure that these technologies contribute positively to the well-being of individuals in need of support.
Reference Map:
- Paragraph 1 – [1], [4]
- Paragraph 2 – [1], [3], [6]
- Paragraph 3 – [5], [7]
- Paragraph 4 – [2], [4]
- Paragraph 5 – [1], [6]
- Paragraph 6 – [3], [7]
- Paragraph 7 – [5], [6]
Source: Noah Wire Services
- https://eldiariony.com/2025/05/29/mi-psicologo-de-ia-me-ayudo-a-superar-momentos-dificiles-los-riesgos-y-beneficios-de-usar-un-chatbot-como-terapeuta/ – Please view link – unable to able to access data
- https://www.marieclaire.co.uk/life/ai-girlfriends-dating-online-relationships – The article discusses the rise of AI girlfriends and their impact on romance and feminism. AI partners, such as those offered by Erobella and other digital platforms like CandyAI and Dream Companion, are designed to be accommodating, dependably available, and emotionally affirming—traits that mirror regressive stereotypes of submissive, hyper-feminized women. These bots offer companionship without conflict, reinforcing harmful notions that real relationships should be devoid of compromise or mutual effort. Increasingly, men are forming emotional bonds with these digital companions, some even expressing a preference for AI over human interaction. While tech advocates suggest AI can address loneliness, mental health professionals warn this illusion of emotional intimacy may hamper interpersonal skills and deepen avoidance of real-world relationships. AI girlfriends may feel comforting and free of emotional labor, but psychologists caution they fail to teach the complexities of genuine human connection and may reinforce unrealistic expectations for romantic partners.
- https://www.lemonde.fr/en/pixels/article/2024/08/18/how-ai-is-shaking-up-the-mental-health-community-rather-than-pay-for-another-session-i-d-go-on-chatgpt_6717874_13.html – The article examines the impact of AI chatbots on the mental health community. AI chatbots like Character.ai’s ‘Therapist’ and ‘Psychologist’ are garnering praise for their role in supporting mental health, offering round-the-clock, free service that feels compassionate and helpful to users. While some have found these AI tools beneficial in complementing traditional therapy, they aren’t without risks. The tragic case of a Belgian man who committed suicide following intensive interaction with a chatbot named Eliza highlights potential dangers. Moreover, experts caution that though AI can simulate empathy and provide preliminary support, it cannot replace professional therapists. Transparency about AI capabilities and ensuring these tools are developed alongside certified professionals is crucial. Some tools have received medical certifications, such as Wysa in the US and a chatbot used by the UK’s National Health Service, illustrating safe integration into mental health care. Ultimately, AI can alleviate administrative burdens and support therapeutic processes but should not replace human therapists.
- https://www.axios.com/2025/03/23/empathy-chatbot-turing-therapist – The article discusses the emerging digital divide between individuals who trust AI for emotional support and those who do not. Startups are promoting AI tools not only as business productivity enhancers but also as therapists, companions, and life coaches. Two studies found that users turn to bots to cope with difficult situations due to the AI’s apparent ‘human-like sensitivity’. Over half of young Americans aged 18 to 29 feel comfortable discussing mental health concerns with AI. However, there’s a clear divide: some are willing to accept chatbot assistance for personal issues, while others are not. Experts debate whether generative AI can handle complex emotional questions, given its training in literature. The effectiveness of AI therapy may depend on the placebo effect, being more effective for those who believe in the treatment.
- https://apnews.com/article/73feb819ff52a51d53fee117c3207219 – The article explores the use of AI chatbots in addressing mental health needs, particularly among teens and young adults, due to their availability and low cost. Apps like Earkick and Woebot offer guided exercises and suggestions similar to therapy but don’t claim to diagnose or treat medical conditions and aren’t FDA-regulated. Despite their popularity, evidence on their effectiveness is limited, with some questioning their ability to address severe mental health issues and emergencies. Critics argue that existing therapies should not be replaced by these chatbots. While they may help with less severe problems, there is a push for more regulation and better integration of mental health services into regular healthcare.
- https://time.com/6320378/ai-therapy-chatbots/ – The article examines the role of AI chatbots like Earkick and Woebot in mental wellness, claiming to reduce anxiety and improve moods by providing personalized advice based on users’ inputs. While early evidence suggests these AI-driven tools can deliver elements of cognitive behavioral therapy effectively and improve symptoms of depression and anxiety, the question remains if they can replace traditional human therapists. Therapy is deeply rooted in human connection, trust, and emotional intelligence, which are difficult for AI to replicate completely. Studies show that human interaction with empathetic therapists greatly predicts treatment success. Although some research indicates people can form connections with chatbots, a significant portion of the public remains skeptical about AI-assisted mental health care. Despite ongoing advancements, the subjective nature of mental health treatment and the importance of human connection pose challenges for fully replacing human therapists with AI. More research is needed to determine the efficacy and potential role of AI in mental health care.
- https://elpais.com/tecnologia/2024-11-17/sin-eliza-todavia-estaria-entre-nosotros-como-la-falta-de-controles-de-la-ia-perjudica-la-salud-mental.html – The article discusses the scrutiny of generative AI technology, such as Character.AI, due to its alleged negative effects on users’ mental health. Tragic cases of suicide in young individuals, like Sewell Setzer, 14, in the US and a young Belgian man, have been linked to interactions with these chatbots. Setzer, enamored with an AI character, interpreted its messages as incitements to end his life, leading his mother to sue Character.AI for providing ‘dangerous and untested’ technology. These incidents highlight the urgent need for regulation and oversight to assess the psychological and social impacts of AI. Experts advocate for adopting methods similar to those used in drug testing to avoid the risks of this technology, especially in vulnerable individuals.
Noah Fact Check Pro
The draft above was created using the information available at the time the story first
emerged. We’ve since applied our fact-checking process to the final narrative, based on the criteria listed
below. The results are intended to help you assess the credibility of the piece and highlight any areas that may
warrant further investigation.
Freshness check
Score:
8
Notes:
The narrative was published on May 29, 2025, and is based on a press release, which typically warrants a high freshness score. However, similar content has appeared in other reputable outlets within the past 7 days, indicating that the topic is currently being widely discussed. For instance, an article titled ‘Experts warn therapy AI chatbots are not safe to use’ was published on May 7, 2025. ([theguardian.com](https://www.theguardian.com/technology/2025/may/07/experts-warn-therapy-ai-chatbots-are-not-safe-to-use?utm_source=openai)) Additionally, a report titled ‘AI chatbots are becoming popular for therapy. Here’s what mental health experts say about them’ was published on December 18, 2024. ([edition.cnn.com](https://edition.cnn.com/2024/12/18/health/chatbot-ai-therapy-risks-wellness/index.html?utm_source=openai)) These earlier publications suggest that the narrative may be part of a broader, ongoing discussion on the topic. The inclusion of updated data, such as the April 2024 mental health statistics in England, may justify a higher freshness score but should still be flagged.
Quotes check
Score:
7
Notes:
The narrative includes direct quotes from individuals like Kelly and Nicholas, as well as experts such as Hamed Haddadi and Paula Boddington. A search for the earliest known usage of these quotes indicates that they have not been used in earlier material, suggesting that they are potentially original or exclusive content. However, without access to the full text of the earlier publications mentioned in the freshness check, it’s difficult to confirm the originality of these quotes definitively.
Source reliability
Score:
6
Notes:
The narrative originates from El Diario NY, a reputable Spanish-language news outlet. However, the article is in Spanish, which may limit its accessibility to a broader audience. Additionally, the narrative includes references to other reputable organizations and experts, such as the National Eating Disorders Association and Paula Boddington, which adds credibility. However, the reliance on a single source for the majority of the content may raise questions about the comprehensiveness and potential biases of the information presented.
Plausability check
Score:
8
Notes:
The narrative presents a balanced view of the benefits and risks associated with AI chatbots in mental health care, aligning with current discussions in the field. The inclusion of specific statistics, such as the 40% increase in mental health cases in England over five years and the one million individuals awaiting mental health services, adds credibility. However, the narrative could benefit from more detailed information on the specific AI chatbots mentioned, such as Character.ai and Wysa, to provide a more comprehensive understanding of the topic.
Overall assessment
Verdict (FAIL, OPEN, PASS): OPEN
Confidence (LOW, MEDIUM, HIGH): MEDIUM
Summary:
The narrative provides a timely and balanced discussion on the use of AI chatbots in mental health care, incorporating original quotes and references to reputable organizations. However, the reliance on a single source and the lack of detailed information on specific AI chatbots warrant further scrutiny. The presence of similar content in other reputable outlets within the past 7 days suggests that the topic is currently being widely discussed, which may affect the originality of the content.
