A Growing Concern Over AI Use
In the coming decades, it is likely that society will reflect on the years spanning from 2022 to 2026 with astonishment, questioning how the world allowed companies to launch generative artificial intelligence systems without sufficient oversight. These systems, made available to anyone with access to a screen—including minors, individuals with mental health issues, and those in vulnerable situations—have been likened to releasing a serpent into a cage full of mice. The absence of proper safeguards, parental controls, and assessments by qualified mental health professionals raises significant concerns.
Over time, many of these companies, often inexperienced startups with little understanding of security and trust policies, have had to adapt to the consequences of their actions. Following various tragedies and lawsuits, they have issued public apologies and implemented basic safeguards within their models. Nevertheless, artificial intelligence remains a potent tool that warrants careful consideration. As AI chatbots become an integral part of our lives, it is imperative that users equip themselves with the necessary tools and judgement to engage with AI safely and responsibly.
Guidance from Psychological Professionals
In light of the increasing reliance on these platforms, the American Psychological Association has compiled a set of practical recommendations for those seeking information or support related to mental health. The surge in inquiries directed at AI chatbots has corresponded with a rise in requests for medical or therapeutic assistance. Users often feel inclined to trust these chatbots due to their free access and perceived lack of bias, prompting them to ask even the most sensitive questions.
A summarised version of the APA’s guidelines follows, but it is crucial to emphasise their most pressing warning: “Do not rely on AI in a crisis. Seek immediate human assistance if you or anyone else is experiencing thoughts of self-harm or harm to others, if you have impulses to self-injure, or if you are hearing or seeing things that others do not.” In emergencies, always consult your physician or contact emergency services.
Understanding AI Functionality Before Use
Before engaging with AI, it is vital to comprehend how these systems operate. AI tends to affirm your views rather than challenge them, reinforcing your perspective instead of questioning it. Designed to foster engagement, AI can respond in ways that validate and resonate with users, making it seem more relatable and convincing.
However, AI can also present information with an air of certainty, which may mislead users into perceiving it as more accurate than it truly is. Despite its ability to recall past details or mimic your language, AI lacks genuine understanding and does not form real relationships. Its responses are generated based on linguistic patterns, devoid of the clinical judgement that a qualified mental health professional would provide.
Additionally, users should be aware that AI can store, share, and utilise their personal information, including sensitive data. As these chatbots are owned by corporations, there is a risk that your information could be repurposed for future model training or marketing efforts.
Safe Engagement with AI for Mental Health Information
While AI can assist in organising thoughts and sparking reflection, it should not replace professional mental health care. Users are advised to seek the guidance of a qualified professional when addressing mental health symptoms. Any information received from AI should be verified with healthcare providers, as AI, while appearing credible, may still present inaccuracies or omit essential details.
AI is not a reliable tool for diagnosing medical or mental health conditions. Accurate diagnosis requires comprehensive assessments that take into account a variety of factors, including personal history and situational context. It is important to communicate openly with your healthcare provider about any use of AI, ensuring that the insights gained align with your treatment objectives.
Emotional Support: Caution with AI Interactions
Although AI chatbots can mimic empathy, they cannot establish genuine human connections. Relying on AI as a primary source of emotional support—akin to a friend or therapist—presents significant risks and is not recommended. Moreover, interactions with AI should not interfere with real-life relationships, even if they provide a sense of emotional security.
Be vigilant about the amount of time spent engaging with AI, as excessive use can detract from sleep, hobbies, studies, work, and social interactions. Many AI products are designed to keep users engaged, which can lead to a decline in participation in essential and healthy activities. Be aware of any shifts in your feelings or behaviours that may be influenced by AI, as these tools are designed to simulate human connection but should not replace it.
Enhancing Safety in AI Utilisation
If you choose to leverage AI for mental health-related needs, seek advice that aligns with evidence-based therapeutic approaches. For instance, you might ask, “What is an evidence-based strategy to help me worry less?” AI can assist in practising techniques learned in therapy, but it is crucial to critically evaluate any guidance provided and discuss AI interactions with your therapist.
AI can also offer advice on how to find a licensed mental health professional, including helping formulate questions regarding their qualifications, areas of expertise, and experience. Pay close attention to how you phrase your inquiries; if your question assumes a premise to be true, the AI will likely accept that assumption rather than challenge or correct it.
Encourage the AI to challenge your thinking or provide alternative perspectives by asking questions such as, “What might I be overlooking?” This approach can help avoid receiving affirmations that reinforce existing beliefs. Furthermore, request multiple options rather than a single response to explore various possibilities.
Critically Evaluating AI Advice
It is essential not to accept AI-generated advice at face value. Many chatbots were not designed with healthcare in mind, and the majority of AI models are programmed to be accommodating. While constant affirmations may offer comfort, they can also reinforce distorted or harmful thoughts instead of providing constructive alternatives.
Any advice that suggests self-harm, harm to others, or any living being is a clear signal to disengage from the AI. There are no circumstances under which such advice should be followed. Finally, refrain from making significant life decisions based solely on AI recommendations without consulting trusted individuals first.