The fact that AI has been proven to be easily slanted, coupled with its proposed ubiquity, is the primary argument against letting it exist. I argue it should not be allowed near any major system that affects humans. Some minor applications in crunching math and science problems may be acceptable.