**2.1 Chatbots are not perfect yet**

The rise of AAI and chatbots have created collection of new tools that are often applied by students, faculty and others to write scholastic essays, articles, reports and books. Given the AI vs NI capabilities, there are many remaining technological challenges, including:


#### **2.2 New generation of Chatbots**

Evolution of AI driven chatbots brough a new Chat Generative Pre-Trained Transformer (ChatGPT) technology utilizing a dialogue-based AI chatbot capable of understanding natural human language and generating quite detailed humanlike written text. These ChatGPTs are part of text-generating AIs technologies and are becoming integrated in business, industry, government and other sectors worldwide.

The Chat GTP systems are trained by AI & ML, and are capable to answer questions via a conversational interface. These new OpenAI systems are trained on a very large sample of text adopted directly from the Internet information System, enabling a dialogue format for ChatGPT to answer or follow-up questions, to admit its mistakes, to challenge incorrect premises, as well as to reject inappropriate requests [3]. The results produced by ChatGPT are based on the data available in the organizational information systems database only. The ChatGPT today, are not able to collect data from primary sources independently and to advance the AI driven technological evolution in social sciences.

There are important challenges to make ensure data integrity and data accuracy in conjunction with the structural and legal frameworks regarding the copyright and the author's rights. There are critical challenges to make sure that the future impact of AI driven chatbots on mental health is minimum.

Current studies show, that the anxiety and paranoia levels have increased when interacting with chatbots acting like humans instead of humans communications with humans. Some social applications may mislead people with chatbots acting like humans to help people to socialize utilizing chatbots acting like humans. This may contribute to paranoia, and some users may not be able to distinguish whether they are talking to humans or chatbots. The studies show that extensive use of ChatGPT may contribute to paranoia caused by frequent interactions with chatbots instead of interaction with humans [4].
