The primary factor in determining the quality of a review is not the maturity of the reviewer but in my opinion the time that is invested into reviewing the paper. AI research papers are not so complex; generally undergraduates that I have worked with are able to fully understand a paper when given sufficient time. By contrast, the average review you can expect to get at a conference will be from a reviewer who has spent maybe 30 minutes on your paper.

I would encourage you to read the papers that you have been assigned in depth and try to give good reviews. Assuming that there is some discussion period where you can see other reviews on the same paper, you will get some indirect signal on the quality of your review.