A background note can be accessed here: e-Courts Phase III Expands Digital Filing, Remote Hearings and Connected Justice Systems
E-Courts Phase III expands e-filing, virtual hearings and digital access to court records. To what extent can these changes alter the cost and accessibility of justice for litigants, lawyers and witnesses, particularly for those who face geographical, financial or technological barriers?
According to DAKSH’s Access to Justice survey (2015-16), legal fees, lost wages, and foregone business for litigants were estimated to exceed Rs 80,000 crore that year, nearly 0.5% of India’s then-GDP. Programmes under eCourts Phase III can substantially reduce these direct and opportunity costs: virtual hearings and livestreaming can reduce travel and missed workdays for litigants and witnesses, while digital record-keeping through e-filing, digital signatures, and online certified copies can eliminate manual work, uncertainties, and delays.
But these digital processes do not automatically deliver unconditional benefits. If these systems are not tested with affected stakeholders, existing digital divides may be exacerbated. A quick push to digital-only methods can foster exclusion, resistance, and unequal access to justice, particularly for communities operating under customary law, economically disadvantaged litigants, or those most comfortable in languages that common technologies do not readily support. Transition management is key, including providing assistance through e-Sewa Kendras, retaining physical processes as an option, and co-creating systems with judicial officers, advocates, and litigants. Digital delivery might lower costs, but simplicity, transparency, and accountability must be intentional guiding principles.
Phase III seeks to connect courts with police, prisons, prosecution and forensic institutions through interoperable justice systems. Can integrating information across these institutions improve accountability across the criminal justice chain, rather than simply make information move faster between agencies?
Many lacunae in accountability are effectively failures of predictability or of closing task loops: an undertrial remaining in jail after bail is granted, forensic backlogs causing delays, or repeated rescheduling of the physical production of an accused from prison. Though justice institutions collect plenty of data, the key challenge is making it available to the right decision-maker at the time decisions are taken. This calls for action-oriented dashboards, recorded reasons for extensions or rescheduling that can be analysed at scale, and time-calibrated nudges. These can guide compliance with statutory or judicial mandates, such as Section 479(1) of the Bharatiya Nagarik Suraksha Sanhita, 2023 (BNSS) which mandates granting of bail to undertrials held for prolonged periods.
The challenge of inter-institutional coordination lies not just in information management, but also the willingness of various actors to utilise data for decision-making. “Shared data and connected systems” are the solution insofar as lack of top-of-mind reminders was a main factor behind accountability gaps. Courts, police, prisons, and forensic labs differ in data standards and cultures that determine how effectively information can change institutional behaviours. Building real interoperability is less about database-linking than about envisioning a secure, reliable, shared decision-making model, drawing on lessons already learned through other public systems.
The programme introduces AI-enabled capabilities such as transcription, translation, legal research and document analysis into judicial processes. Where should India draw the boundary between using AI as an assistive technology and allowing it to influence judicial reasoning or decisions?
India’s approach is clear: AI must remain assistive, with certain use cases prohibited and the rest carrying mandatory human-in-the-loop obligations. The judicial officer, as the signatory, remains responsible for the integrity of the decision, consistent with the European Union Artificial Intelligence Act (EU AI Act, 2024) and United Nations Educational, Scientific and Cultural Organization (UNESCO) guidelines (2025).
AI is envisioned to achieve three objectives: efficiency through automating rule-based case management and administrative tasks; consistency through better fact extraction and updated research databases, making the basis of judicial reasoning more legible and open to questioning; and accessibility through chatbots, translations and plain-language explanations. Policies curtail unrestrained AI use to protect judicial integrity.
However, the boundary must be defined not by specific functions, but the possibility of harm. Every case weighs competing claims and makes value judgments about rights, a subjective, principle-driven, human exercise. Even relatively technical matters, such as motor accident claims require reasoning beyond the black letter of the law.
Thus, judicial cognition requires the human-in-the-loop to be informed, sincere, and thorough. Judicial officers must verify sources, remain sceptical of AI outputs, and be clear about their own reasoning rather than depend on a tool that could produce varying recommendations each time it is used. Otherwise, judges risk automation bias and the erosion of junior lawyers’ expertise as entry-level work is automated away.