JoSAA 2026: How CBSE's OSM Errors Trapped JEE Qualifiers Below the 75% Threshold

The Unlikely Paradox of 2026's Board Season

There is a particular cruelty in failing the easier exam after passing the harder one. In June 2026, that paradox became a material reality for an unknown but growing number of Indian students. These students had cleared JEE Advanced — widely regarded as one of the most competitive entrance examinations in the world, with a success rate of roughly 2.5 percent among over 1.5 lakh test-takers — only to discover they no longer met the mandatory 75 percent Class 12 board marks threshold required for IIT and NIT admission.

The reason was not poor academic performance. It was a scanning error. Or a blurred page. Or a wrongly attributed answer sheet. Or an evaluator marking on a set of scans that had already been identified as illegible in CBSE's own internal quality logs.

This is the hidden cost of India's first full-scale On-Screen Marking (OSM) deployment, and the JoSAA 2026 admissions cycle has now become the arena where that cost is being tallied.

The 75 Percent Rule: A Requirement That Has No Exceptions

Under JoSAA (Joint Seat Allocation Authority) rules that govern admissions to IITs, NITs, IIITs, and Government Funded Technical Institutes (GFTIs), candidates from the general, OBC-NCL, and EWS categories must secure a minimum of 75 percent aggregate marks in Class 12 board examinations. SC, ST, and PwD students must clear 65 percent.

This requirement was introduced as a filter against students who might perform well in coaching-driven entrance preparation while neglecting school-level conceptual grounding. It is a pedagogically defensible policy. It also assumes that the board marks printed on a student's result sheet are an accurate reflection of their actual performance.

In 2026, that assumption collapsed.

What CBSE's OSM Rollout Produced

CBSE launched its OSM system at unprecedented scale for the 2026 Class 12 evaluation cycle. Approximately 98,66,622 answer books — representing nearly 40 crore scanned pages — were processed through the system. The national pass percentage fell to 85.20 percent, the lowest in seven years, down from 88.39 percent in 2025.

The individual failure modes were numerous and serious:

CBSE engaged cybersecurity experts from IIT Madras and IIT Kanpur after a 19-year-old researcher publicly disclosed that vulnerabilities in the OnMark portal had been reported to CERT-In in February 2026 but remained largely unpatched through the May results cycle.

For students already on the margin of the 75 percent threshold, a single paper's marks being under-credited due to an unchecked page or an illegible scan can be the difference between an IIT seat and none.

IIT Roorkee's Response: Procedural Relief, Not Rule Relaxation

On June 6, 2026 — the same day student protests erupted outside the Ministry of Education in Delhi — IIT Roorkee, acting as the lead JoSAA institute, announced a procedural accommodation. Students currently below the 75 percent threshold would be permitted to:

However, final admission remains contingent on the student demonstrating the 75 percent mark before the last reporting date. The conditions under which this can happen include the student's original mark sheet being correct (i.e., the board has corrected a scanning or marking error through the revaluation process), or the student clearing their marks through re-evaluation, supplementary examination performance, or other formal board review.

IIT Roorkee was explicit: this is not a waiver of the eligibility rule. The rule applies equally to students from 36 different boards. Systemic relaxation was described as "not possible" at this stage.

A Racing Clock

The situation places affected students in an extraordinarily compressed timeline:

A student whose marks are under review due to OSM errors must therefore simultaneously track a CBSE revaluation process — which typically takes four to six weeks — while participating in a seat allocation process whose Round 1 outcome arrives before revaluation results. The conditional allotment mechanism was designed precisely to prevent students from being shut out entirely during this uncertainty, but it does not resolve the underlying marks problem.

What This Design Failure Exposes

The JoSAA 75 percent crisis illustrates a structural risk that exists wherever a downstream process — in this case, elite university admissions — takes the outputs of an upstream evaluation system as accurate without any mechanism to flag or hold on quality exceptions.

A properly designed digital evaluation system addresses this at the source. Before marks are declared, the platform must:

Enforce scan quality gates

Every page of every answer booklet should be processed through automated image quality checks before it reaches an evaluator. Pages failing resolution or legibility thresholds must be flagged for re-scan, not forwarded. CBSE's own data showed that 68,000 answer sheets required a second scan, and over 13,000 still failed to meet digital standards and had to be evaluated manually — a process that should have been the exception, not a revelation after results were declared.

Maintain identity integrity at the booklet level

The problem of students receiving answer sheets belonging to others points to a booklet-level identity management failure. Each booklet must carry a machine-readable, cryptographically verified identifier that the system cross-checks against the candidate's barcode at every stage — scanning, stitching, evaluator upload, and final marks compilation.

Provide real-time quality dashboards for administrators

If board administrators had had live visibility into the volume of rejected scans, portal error rates, and evaluator session completions during the marking window, the scale of quality failure would have been apparent in late February or March — not May, after results were declared and JoSAA counselling was imminent.

Generate complete, auditable marks derivation trails

For every question in every answer book, the system should record which evaluator marked it, at what time, with what assigned mark, and whether any flag was raised. This trail makes post-declaration review a matter of hours, not weeks.

The Broader Signal for Institutions

CBSE's experience is not an argument against digital evaluation. It is an argument against deploying large-scale digital evaluation without piloting it, testing it at load, auditing its vendor's track record, and verifying its data integrity mechanisms before the results of 17 lakh students depend on it.

Universities and boards watching the JoSAA 2026 situation unfold would do well to note what was absent from the CBSE rollout: phased deployment, independent security audits completed before production, evaluator training validated through mock cycles, and backup protocols for scan failures. Institutions considering their own OSM implementation have the opportunity to build these safeguards in from the start, before any student's admission is contingent on the output.

---

Related Reading