This paper aims to study the Brazilian public policy for fomenting innovation in Artificial Intelligence (AI), presenting the initial premise (hypothesis), to be inductively tested, that its greatest challenge is related to the endogenization or internalization of the processes of development and production of AI-related technologies in the country. To this end, we analyze data from the patenting of these technologies in Brazil, the most widely used indicator to measure national technological innovation, contrasting them with international data. Considering the diversity of formats that such public policies can take, this work reveals its importance, since it provides an accurate diagnosis of the reality in the AI segment in Brazil, a fundamental subsidy for the formulation of efficient planning.