According to Kaspersky, its antivirus software "automatically uploaded [malware samples] to Kaspersky’s specialists for further analysis." Is this common practice for antivirus software? How do they make sure they're not uploading personal files infected with malware?
I understand that antivirus software initially checks some hash of samples found on an infected computer and checks a database for a match first. I'm asking about the case where they don't find a match and have a potentially "new" malware.