Reforms do not appear out of nowhere. The 2026 report exists because the CLAT has had a question-quality problem for years, and it finally reached the courts. To understand where the exam is heading, it helps to see clearly what was going wrong. The committee read several years of past papers, both timed and at leisure, and grouped the flaws into a handful of recurring patterns.
None of this affects your 2027 prep — the pattern is unchanged. But it explains why the exam is moving the way it is, and why "learn to reason" is safer advice than "memorise more."
Problem 1: answers hiding in plain sight
The most common complaint is simple. In a lot of questions, the answer sat word-for-word in the passage. You did not need to understand anything — just find the matching line. The report calls this reducing comprehension to mere recognition.
Its example from the English section: a passage says the Scottish moors "serve as an ever-present omen of the impending tragedy," and the question asks what role the moors play in Macbeth. The correct answer is that they are an ever-present omen of impending tragedy. That is copy-paste, not comprehension. The same thing showed up in Legal Reasoning, where a question asked what "online defamation" meant and the definition was printed right there in the extract.
Problem 2: more than one defensible answer
A second recurring flaw: questions where two or three options could each be argued for. The report shows one on the Jallianwala Bagh massacre that asks why General Dyer ordered firing on the crowd — and the options (to bring the crowd under control, to scatter it, to teach it a lesson, to enforce martial law) are all plausible readings. A Logical Reasoning sample asks which conclusion follows from a florid sentence about "an engagement with words," where several options are equally arguable. When a question has no single clear answer, it stops measuring anything, and it undermines both the reliability and the fairness of the test.
Problem 3: no correct answer at all
Worse than an ambiguous question is a broken one. The report highlights a Legal Reasoning question about the role of intermediaries in online defamation where every single option was demonstrably wrong. It had to be withdrawn after the exam. When that happens, students who spent time on it were penalised for a mistake that was not theirs. The committee treats this as a distinct and more serious category — questions that are not merely poorly designed but factually or logically incorrect, leaving them literally unanswerable.
Problem 4: knowledge the test claims not to require
The CLAT says it tests aptitude, not prior legal knowledge. But the report found questions that quietly assumed you had studied law. One asked whether "an agreement made without consideration is generally" valid, enforceable, void, or voidable — a straight contract-law question a school-leaver has no reason to know. As one Vice-Chancellor put it, asking students questions from criminal law or contract before they have studied law "is unnecessary and benefits coaching centres more than law schools." Questions like this reward those who paid for coaching and disadvantage everyone else.
Problem 5: the sections all tested the same thing
Step back, and a structural issue appears. The paper is split into five sections meant to test different aptitudes, but in practice most of them tested one thing — a shallow reading comprehension. That defeats the point of having distinct sections. On top of it, consecutive papers overlapped heavily in theme, with topics like juvenile justice, data protection, and corporate consultancy recurring year after year. That repetition rewards students coached on predictable themes rather than genuine aptitude.
The postgraduate paper's own problems
On the PG side, the passages grew alarmingly: from about 5,404 words in 2023 to roughly 8,954 in 2024, and still around 8,413 in 2025. Combined with a vast syllabus and heavy reliance on recall, the paper tested memory and stamina more than reasoning. The report's examples of pure rote include a question asking which judge disagreed with a particular Sabarimala opinion, one asking candidates to translate a Latin maxim, one on when India's first administrative tribunal was set up, and one on when the National Green Tribunal Act was passed.
An honest caveat from the committee
The report is fair about this: the errors are not uniform. Some years the papers are largely clean; other years they are riddled with problems. It is precisely this unpredictability that worries the committee. As one Vice-Chancellor said, only slightly exaggerating, there had "not been a single year where some answers in the official key were not contested or withdrawn." That makes the problem systemic, not incidental — a reason to build a consistent process rather than hope each year's team gets it right.
The CLAT litigation timeline
This was never just an academic complaint. Concerns reached the courts as far back as the Disha Panchal case, where the Supreme Court said it was "not at all satisfied" with how the exam was conducted and questioned the model of rotating oversight between universities.
It came to a head with CLAT 2025. Students filed petitions in several High Courts. A single judge of the Delhi High Court found errors in two questions; the cases were then consolidated at the Delhi High Court, whose division bench ordered four questions withdrawn as erroneous and the rank list revised. On the PG side, the Consortium withdrew five questions and the court struck down two more. On appeal, the Supreme Court recorded its "deep anguish regarding the callous and casual manner in which" questions were being framed, and pointed out that despite earlier directions, nothing had changed in how the exam was designed and run. That judgment is the direct trigger for the Expert Committee.
Why this matters to you
Every one of these flaws has a human cost. Broken questions mean litigation, delayed admissions, and real anxiety for students who did nothing wrong. And they hit students without money for coaching or lawyers the hardest — the exact opposite of what a fair gateway should do.
The lesson for your own prep is quiet but real. The exam is moving away from memorised facts and toward careful reasoning. Facts can be tested badly, withdrawn, or disputed. The ability to read closely and reason from what is in front of you is what every version of this exam is trying to reward — so that is the safest thing to train, whatever year you sit.
Frequently asked questions
Why is the CLAT being reformed?
Because CLAT papers have carried recurring errors for years — answers found verbatim in the passage, ambiguous questions with more than one defensible answer, questions with no correct answer, and questions needing prior legal knowledge — which led to repeated court cases and, in 2025, the Supreme Court expressing "deep anguish" at how questions were framed.
Has the CLAT had wrong questions?
Yes. The 2026 report documents questions with no correct answer (one intermediary-liability question was withdrawn after every option was demonstrably wrong), questions with multiple defensible answers, and answers printed word-for-word in the passage. Courts have ordered several questions withdrawn and rank lists revised.
What did the Supreme Court say about the CLAT?
In a 2025 judgment, the Supreme Court recorded its "deep anguish regarding the callous and casual manner in which" questions were being framed, and noted that despite earlier directions, nothing had changed in how the exam was designed and conducted.
What are the main problems with CLAT questions?
The report groups them into five patterns: answers hidden verbatim in the passage; questions with more than one defensible answer; questions with no correct answer at all; questions assuming prior legal knowledge; and sections that all test the same shallow reading comprehension rather than distinct aptitudes.
Do CLAT questions test memorisation?
Too often, yes — especially in the PG paper, which the report found relied heavily on recall of dates, judge names, and Latin maxims. The reforms aim to shift the exam toward reasoning and away from rote recall.