Some countries and industries are known to have higher risks of corruption or labour abuses. If an AI flags companies from these places more often, is it reflecting real risk or unfairly discriminating against honest businesses? Where should the line be drawn?