The quality of the data used to train and contextualize AI systems is crucial to ensure the reliability and impartiality of the results they generate. Users must use accurate and complete data, that reflects the reality they would like to analyse or model. 

Low-quality, incomplete or biased data can lead to inaccurate, misleading or biased results. It is essential that training data is free of bias, which could be reflected in the results generated by AI, perpetuating stereotypes or discrimination. Ensuring the quality of the context data provided to AI is equally important. 

The prompt, i.e. the input that AI uses to generate new content, must be clear, precise and developed in such a way as to avoid ambiguity or misinterpretation. A vague or poorly formulated prompt can have a negative impact on the quality of results and lead to inconsistent or irrelevant outputs. Users must provide AI with an appropriate context, formulating clear and specific prompts, in order to obtain reliable and high quality results. 

Users must be aware of the possible biases present in Generative AI models and adopt a critical approach to avoid spreading discrimination or prejudice.