It is fairly common to get this observation when conducting interaction effects test in ANOVA.
Since you've found the interaction to be significant, you can interpret that
the variables A and B jointly are able to explain the variation in your response variable.
One important thing to note is that when you've found the interaction to be significant, you have to keep the individual variables in the function.
In this case, A:B is significant so obviously interaction term should be present. In addition to this, you need to keep A and B in the equation as well even if they're not significant.