AI Harm Severity Scales by Category

Note: I have removed the 'Additional Definitions' and 'Other' categories from the CSET taxonomy and changed the 'Detrimental Content' category to 'Toxic or Malicious Content'.