雅思写作 Task 2

As artificial intelligence plays an increasingly prominent role in critical decision-making domains such as hiring, loan approvals, and criminal justice, ensuring algorithmic fairness has become a central issue in public discourse. When algorithms are trained on historical data—which may itself carry societal biases—should we prioritize "outcome fairness," actively adjusting algorithms to correct historical injustices? Or should we uphold "procedural neutrality," where algorithms act solely on verifiable facts and rules, never considering group identity, as the cornerstone of justice? Drawing on concrete examples from specific fields, articulate your position and justify its validity.