Introduction / Theoretical framework:
Artificial Intelligence (AI) is transforming various fields, including education. In particular, generative AI, which creates original content from existing data, has been integrated into the academic environment as a complementary teaching tool. In the context of the Pharmacy Degree at the Complutense University of Madrid, generative AI was implemented in the Analytical Chemistry I course as part of an educational innovation project (INNOVA “Teaching-Innovating-Advancing-Generating: Generative Artificial Intelligence”). The project aimed to enhance the teaching of practical concepts and encourage critical reflection among students regarding the use of these tools.

Objectives:
- Foster independent learning through the use of AI to solve practical exercises.
- Develop critical thinking and problem-solving skills by identifying errors in solutions provided by AI.
- Measure the effectiveness of AI as an educational tool for improving practical learning of titrations and numerical exercises.

Methodology:
Three theory groups were included in the study: A (94 students), A1 (95 students), and B1 (100 students). Two activities were carried out throughout the course, one in the middle of the subject and another at the end. The activities were carried out in groups of 4–6 students. In these activities, students:
1. Detected errors in the exercises that were previously solved with AI, knowing that there were errors but not knowing which ones.
2. Solved new exercises using AI and assessed whether the solutions were correct, identified mistakes, and tracked the number of attempts until the correct solution was reached.
Students filled out two questionnaires (one for each activity) in which data were collected on the number of errors made by AI, the type of error, and the number of attempts to find the correct solution. This data helped measure the impact of AI on students’ learning.

Results and discussion:
The percentage of class attendance and participation in the activities was 70%. Students were asked in the survey about the AI tools they employed. None of the student groups obtained the correct answer from the AI on the first attempt, regardless of the AI employed (Gemini, Copilot, Open AI, etc.). The majority of the errors made by AI were conceptual or numerical mistakes in mathematical operations.

Conclusions:
The project of implementing generative AI in Analytical Chemistry I demonstrated significant potential in improving students' practical learning. Through the proposed activities, students not only improved their problem-solving skills but also developed a deeper critical understanding of the use of technological tools in education. The survey results provided valuable insights into the real impact of these activities on teaching and learning in the course.

Keywords: Generative AI, Analytical Chemistry, Educational Innovation.