Healthy Huskies: The potential of AI in treatment settings

Treatment for mental health can be a difficult road to navigate. As more people struggle with their mental health, the spaces of psychology, treatment and counseling must evolve. For years, mental health providers have relied on “traditional” treatment routes. Avenues such as talk therapy, medication and treatment facilities have kept many patients afloat for decades. While there are still many downsides to these methods of treatment, they are regarded as the most useful in the battle against mental illness. However, less “traditional” methods of treatment have come to light in recent years. Methods such as transcranial magnetic stimulation (TMS) and psychedelic therapy have begun to make their way into the toolboxes of treatment providers, allowing for new breakthroughs. Another method has been discussed recently: the use of Artificial Intelligence (AI) for mental health treatment.  

AI has skyrocketed in popularity recently. AI programs such as ChatGPT can be used for virtually anything, from creating recipes or helping craft study plans. Recently however, there has been discussion in the medical community regarding whether or not AI can be useful in helping to treat patients. Some believe that the use of AI in the medical field could have the potential to cause more harm than good, arguing that AI is not an adequate tool to aid healthcare professionals. As advanced as the technology is, it is no replacement for an actual healthcare worker. However, some also argue that AI can be a useful tool for healthcare professionals by aiding in providing resources and analyzing data to provide a diagnosis.  

When it comes to mental health, the argument is just as divided. In America alone, about 44 million individuals are living with a mental illness. Due to the COVID-19 pandemic, these numbers increased in recent years, with more people struggling now than ever. Treatment for mental illness traditionally follows routes such as cognitive behavioral therapy, medication management and various levels of treatment programs if needed. All these avenues rely on the use of mental health workers. However, the use of AI in treatment settings may become more commonplace as time goes on. Institutions such as the World Health Organization suggest that AI has the potential to be used in analyzing patient data and progress notes to come to a final diagnosis. The technology also has the potential to help psychiatrists manage and prescribe medication by taking into account symptoms and health data, in order to find different medications and dosages that may help the patient. Not to mention AI can also provide many crisis resources to anyone struggling.  

Many argue that AI is not a safe replacement or supplement for actual mental health care. Psychiatry and therapy rely heavily on human-to-human connection, and so using AI for these treatments may not be a smart idea. There have been arguments that AI can be used as a sort of “therapist” —- helping users with issues such as anxiety or sadness. However, if you are struggling with a serious mental health concern, such as depression or issues with eating for example, AI is not the place to turn to. As advanced as the technology is becoming, artificial intelligence is not equipped to handle serious mental health issues. Additionally, allowing AI to input patient data into an online system could have the potential to violate many HIPAA patient privacy laws. Many mental health outlets have also remarked on the system’s potential to adhere to what is called “algorithmic bias.” Algorithmic bias is described as a system not considering differences in user experience such as disability, race, gender, sexuality and so on. AI treats each user the same, not noticing important differences between them. This can create a large bias that may gravely harm a patient rather than help them.  

In conclusion, there are many different viewpoints when it comes to discussing the potential use of AI in mental health treatment. AI can provide interim resources to many different users and could be used in making diagnosis or managing medications. However, there are also many potential flaws when thinking about the use of AI in the field of mental health. AI has not advanced enough to provide unbiased and confidential counseling, as AI has a hard time grappling with emotion and differences in patients. Additionally, uploading confidential patient data into an AI system has the potential to violate many patient privacy laws. While AI could be utilized as a helpful tool for mental health providers, it is no replacement for help and support from an actual person.