As you have sample data, you are testing the inference that the population parameter is significantly different from zero. The confidence interval gives you a range of values taking into account or reflecting how much error you have I this estimate. So if the confidence interval contains the zero you cannot reject the hypothesis that it the population parameter is equal to zero. Remember your parameter estimate is simply a mean not the actual population parameter which is unobservable
It means that the confidence interval doesn’t contain Zero between the upper value and lower value. For example Ll= 0.3 and Ul= 0.6 if you draw a confidence interval scale and try to position these two values it wont include Zero in between. This means that the path is significant.