Harvard researchers reveal a critical paradox in artificial intelligence development: rather than eliminating human prejudices, AI systems are actively perpetuating and amplifying them. Professor Mahzarin Banaji's findings demonstrate that machine learning algorithms inherit biases from their training data and creators, resulting in discriminatory outcomes across hiring recommendations, salary predictions, and career guidance. This phenomenon raises urgent questions about algorithmic accountability and the need for ethical oversight in AI implementation across industries.

Summary auto-generated by AI from the original publisher's content. Editorial standards.