{"id":8898,"date":"2026-05-16T15:20:06","date_gmt":"2026-05-16T15:20:06","guid":{"rendered":"https:\/\/www.wendyzhou.se\/blog\/?p=8898"},"modified":"2026-05-16T15:20:11","modified_gmt":"2026-05-16T15:20:11","slug":"using-ai-for-self-reflection-benefits-risks-and-how-to-get-started","status":"publish","type":"post","link":"https:\/\/www.wendyzhou.se\/blog\/using-ai-for-self-reflection-benefits-risks-and-how-to-get-started\/","title":{"rendered":"Using AI for Self-Reflection: Benefits, Risks, and How to Get Started"},"content":{"rendered":"\n

AI tools like ChatGPT and Claude are increasingly being used for self-reflection, mental health support, and digital journaling \u2014 but is it actually safe and helpful?<\/strong> When you feel bored, sad, anxious, or indecisive about a life decision, have you ever turned to AI to talk about it? <\/p>\n\n\n\n

Modern AI tools such as ChatGPT and Claude offer a conversation-based interface, allowing users to interact with them as if they were talking to a friend or a therapist. Many recent research initiatives have begun to explore the ways AI is used for emotional well-being and personal growth. In this article, we’re going to look at the benefits, risks, and how you can utilize AI for guided self-reflections.<\/p>\n\n\n\n

What Are LLMs? (The AI Behind ChatGPT and Claude)<\/h2>\n\n\n\n

Artificial Intelligence (AI) is a type of technology that simulates human intelligence. Modern AI has become highly advanced in particular language abilities. We call these types of AI models Large Language Models (LLMs). As the name suggests, LLMs are trained on large amounts of language data and are very good at handling natural language. Tools such as ChatGPT and Claude are based on LLM models.<\/p>\n\n\n\n

When you are chatting with ChatGPT and Claude, you are actually interacting with an underlying LLM.<\/p>\n\n\n\n

Using AI for Self-Reflection and Mental Health<\/h2>\n\n\n\n

With ChatGPT and Claude, you can talk to them as if they were another human being. You can ask AI questions, share things, and ask for advice. These abilities create an emotional connection, similar to how we would feel if we confided or had a conversation with a person.<\/p>\n\n\n\n

In addition, AI can feel safer than talking to a real person, since it can’t really see you or do any harm to you in person. It can feel anonymous. It can even feel safer than talking to a therapist, because this way, you don’t need to expose yourself to anyone else, yet still get the support and validation you might need.<\/p>\n\n\n\n

Using AI to Guide Self-Reflections<\/h2>\n\n\n\n

You can use AI to help you go through your problems and concerns. AI can ask you follow-up questions, guide your reflections and thoughts towards a more positive framing, and stop you from ruminating. <\/p>\n\n\n\n

When self-reflecting with AI, you can ask it for advice and ask it to consider multiple perspectives and possible trajectories about the future. You can also ask it to research the internet for you, to provide you with informed answers.<\/p>\n\n\n\n

One tip: If you want honest advice, ask AI to be realistic, to consider multiple perspectives, and not to provide you with an answer simply to appease you.<\/p>\n\n\n\n

Given this, I think there’s great potential in how AI can be used for digital journaling, to improve your mental well-being, and to help you gather data to make decisions.<\/p>\n\n\n\n

Risks of Using AI as a Therapist<\/h2>\n\n\n\n

There are, of course, risks with using AI for self-reflection. For example, your personal data and concerns are very likely to be used to train the AI model, and who knows who’ll have access to that in the future. This is your most personal thoughts and reflections we’re talking about, and it should be concerning. <\/p>\n\n\n\n

In addition, AI is not a trained therapist. We cannot be sure that the advice it gives you and the help that you experience from it are truly good for you. It might feel like it is, but there have been many cases where AI has given poor advice to people.<\/p>\n\n\n\n

The tech companies are trying to improve the models, placing safeguards, and aiming to improve the models to become more ethical. However, they cannot be certain about the potentially harmful effects of using AI for mental-health concerns or difficult life decisions.<\/p>\n\n\n\n

If you’re interested in AI journaling, mental well-being, or self-reflection, check out these further readings:<\/p>\n\n\n\n