The study investigates whether Large Language Models (LLMs) can generate empathic responses to support seekers in various everyday situations. Two studies were conducted with a total of 192 participants, who rated responses generated by LLMs (GPT4 Turbo, Llama2, and Mistral) and human-written responses on their perceived empathy. The results showed that LLM-generated responses were consistently rated as more empathic than human responses. Linguistic analyses revealed distinct "styles" of responses from different LLMs, such as Llama2 being more casual and using more emojis, while Mistral responses contained more apologies and acknowledged negative emotions. The study highlights the potential of LLMs to enhance human peer support, particularly in contexts where empathy is crucial. However, ethical concerns about the inherent deception and biases in AI-generated empathy are also discussed.The study investigates whether Large Language Models (LLMs) can generate empathic responses to support seekers in various everyday situations. Two studies were conducted with a total of 192 participants, who rated responses generated by LLMs (GPT4 Turbo, Llama2, and Mistral) and human-written responses on their perceived empathy. The results showed that LLM-generated responses were consistently rated as more empathic than human responses. Linguistic analyses revealed distinct "styles" of responses from different LLMs, such as Llama2 being more casual and using more emojis, while Mistral responses contained more apologies and acknowledged negative emotions. The study highlights the potential of LLMs to enhance human peer support, particularly in contexts where empathy is crucial. However, ethical concerns about the inherent deception and biases in AI-generated empathy are also discussed.