JUSTICE BEHIND THE BLACK BOX: ALGORITHMIC ACCOUNTABILITY AND THE LIMITS OF PAKISTAN’S LEGAL FRAMEWORKS

Abstract

The adoption of AI in the judiciary of Pakistan has shown numerous efforts to increase the efficiency of the judicial process while retaining the human monopoly on judicial decision-making. However, the increasing use of AI introduces major risks, such as biased programming, black-box reasoning, and technical hallucinations. Existing laws are completely unequipped to handle these issues because there is currently no legal framework that governs how AI is used in judicial matters or daily life. Even existing laws like the Prevention of Electronic Crimes Act (PECA) 2016, which deals with online and electronic crimes, lack specific requirements for algorithmic accountability, creating a regulatory void regarding biased programming and errors caused by using AI. Establishing robust AI accountability is vital for maintaining the moral authority of the court, judicial fairness, and public trust, ensuring that justice remains a product of human conscience rather than opaque machine logic. This study examines Supreme Court Precedents and national policies to see if they are effective in addressing these challenges. To help strengthen legal control, this study proposes a “Judicial AI Accountability Framework” which identifies liabilities, requires transparency regarding the use of AI in courts, and introduces judicial audit trails—unchangeable forensic digital records that document all judge-AI interactions to increase traceability. This helps to turn policy promises into robust legal protections and keep judicial decisions credible, even as automation increases. This study uses doctrinal and qualitative methodology that relies on statutes, precedents, policy documents, news articles, and academic literature. Ultimately, this research contributes to creating effective regulatory mechanisms for ethical and responsible use of AI in the Pakistani Judiciary. This research finds that existing statutes, policy, and precedents provide no operational accountability for AI-assisted adjudication. It recommends codifying the proposed Judicial AI Accountability Framework through a dedicated statute and an independent oversight authority to translate promises into binding legal protections.