This paper presents a discussion on Enactive Artificial Intelligence about the practical evolution of this field. It has been mostly discussed theoretically, but few are the works that try to implement it. In 2009 were published two paradigmatic works of the field, where the authors state requirements for Enactive Artificial Intelligence and for agency, in an enactive perspective. We take those works and their requirements to analyze other papers that claim to have a practical experiment on Enactive Artificial Intelligence. Although they are a few in number, they made some progress in respect to the requirements. The agency conditions were not really a problem, for almost all works were able to achieve them. On the other hand the Enactive Artificial Intelligence design principles proved to be hard. Only one work was able to implement one of them and none for the other, although one work seems to have a potential to advance in this principle. We conclude that the combination of the works may be the future of the Enactive Artificial Intelligence field.