How to calculate Confidence Interval from correlation latent factors in CFA?

If want to calculate the 95% CI of the interfactor correlations (correlations between latent factors) in a CFA, I believe I do this by multiplying the standard error by 1.96 and add/ subtract this from the correlation? However, AMOS only gives standard errors of the covariances, not of the correlations. How do I calculate standard errors for the correlations? or how do I calculate the 95% CI of the interfactor correlations? I hope anyone can help me with this question.
Perhaps just bootstrap 2500 samples or so and use the bootstrapped standard error that AMOS gives you to build the confidence intervals. I imagine that it's likely to give you a value quite close to the normal ML derived standard error of the statistic. If they were to be different, I'd imagine it's due to some normality issues in your sample which I've seen folks suggest dealing with via the bootstrapping feature in AMOS.