"Automation in evaluation systems does not automatically guarantee objectivity or fairness." Critically examine in the context of recent controversies in Indian public examinations.

Automation in assessment promises speed, uniformity and freedom from examiner bias. Yet the 2026 CBSE On-Screen Marking (OSM) episode shows that technology only relocates discretion — fairness depends on the institutional safeguards built around the machine, not the machine itself.

Where automation does strengthen objectivity

• Standardisation: OSM replaced physical spot-evaluation camps, with CBSE citing "efficiency and transparency" as the goal for Class XII evaluation from 2026 [1].
• Anonymity and auditability: digital scripts mask candidate identity and leave time-stamped trails of marks entered, enabling monitoring of examiner drift.
• Speed and scale: near-a-crore scripts were digitised, cutting the logistics of transporting bundles to examiners' homes.

Why automation alone did not guarantee fairness

• Garbage-in problem: blurred scans, missing pages and mismatched sheets meant examiners marked incomplete evidence — an input failure no algorithm corrects.
• Rushed rollout: introduced by a February 2026 circular for the same exam cycle [1], with trial runs hit by portal and login glitches; teachers received limited training.
• Redressal deficit: students who cleared the board but fell below entrance cut-offs found courses already underway, prompting the Supreme Court to ask CBSE to "find out how you can protect their interest" [2].
• Transparency dilution: curtailed post-result access sits uneasily with CBSE v. Aditya Bandopadhyay (2011), which held evaluated answer-books to be "information" open to inspection under the RTI Act [3].
• Accountability gap: outsourcing core evaluation to a private portal operator diffuses responsibility for errors.

Assessment: automation is procedurally neutral but substantively only as fair as its design, piloting and grievance architecture — a lesson the Radhakrishnan Committee on NTA reform had already drawn for examination integrity and data security [4].

The way forward lies in phased pilots before full rollout, mandatory scan-quality verification, a statutory right to view one's evaluated script, time-bound and low-cost redressal, and independent audit of vendor systems. Digital evaluation must be an instrument of the transparency it promises, not a substitute for it.