Thank you for testing... I optimized the algorithms quite a bit today, so they have changed dramatically. As I mentioned before, they were a little aggressive, but I think we have them right were we want them now. I updated the stand alone VoodooAi software, and I will post a link soon. It has the brand new algorithms and adjustable VoodooAi sensitivity, so it will be pretty cool... you guys will be able to help me figure out what level to set it to by default. Keep in mind, we are now working with completely different models, so it is going to react much different from before. I have not tested the new models thoroughly, but as far as I can tell so far, 100% is pretty close to optimum. BTW, some files like dnsjumper are going to test high, no matter what we do. I can send you a link to that malware pack if you email or pm me. If you download the new standalone VoodooAi, it will be super easy to test anything you would like. One last thing, please keep in mind... the raw data for VoodooAi and VoodooShield are going to be the same (like if you click on Details in VS / test with VoodooAi).... but the composite VoodooAi score for VS is adjusted slightly when there are blacklist hits... and if there are a high number of blacklist hits, then the composite VoodooAi score is adjusted a little more. The whole idea being that the purpose of Ai is to think for the user, so they do not have to... basically, it helps them to make the correct decision whether to allow or block something. So a while back, I figured that the blacklist scan should be one of the "features" of VoodooAi. I know I mentioned this before, but I just wanted to remind everyone . Thank you!