Mukesh Singh v. State (Narcotic Branch of Delhi): informant and investigator can be the same

The question referred to the Constitution Bench

For a period after 2018, NDPS prosecutions across the country carried a structural vulnerability. In Mohan Lal v. State of Punjab (2018) 17 SCC 627 a three-judge bench had held that where the first informant and the investigating officer are one and the same person, a "reasonable apprehension" of bias arises, the fairness of the investigation is compromised, the trial stands vitiated, and the accused is entitled to acquittal on that ground alone. The rule operated as a per se disqualification: it did not ask whether the accused had suffered any actual prejudice.

Because narcotics work is frequently a single-officer affair — the officer who receives secret information and effects the search and seizure is often the officer who then investigates — Mohan Lal threatened a large class of convictions. Later benches doubted its correctness, and a two-judge bench in Mukesh Singh referred the point to a larger bench. A five-judge Constitution Bench of Justices Arun Mishra, Indira Banerjee, Vineet Saran, M.R. Shah and S. Ravindra Bhat was constituted to settle the law. Justice M.R. Shah authored the unanimous judgment delivered on 31 August 2020.

What the Bench decided

The Constitution Bench answered the reference squarely: the informant being the investigator does not, without more, taint the investigation or entitle the accused to acquittal. Whether an investigation is in fact unfair or biased is a question of evidence to be resolved case by case, not a presumption to be applied mechanically.

The reasoning proceeds in three steps. First, the NDPS Act does not bar the informant from investigating the case; there is no statutory prohibition to that effect, and the courts cannot read one in. Second, the credibility of an official witness cannot be discarded merely because he is also the investigator — an officer is presumed to act honestly and in accordance with law, and the presumption is not displaced by the coincidence of roles. Third, and most importantly, bias is a matter of proof. An apprehension is not evidence. The accused who alleges that the overlap produced an unfair investigation must demonstrate it on the record; the burden does not shift to the prosecution to disprove a presumed bias.

The overruling of Mohan Lal

The Bench was direct about the fate of the earlier rule. It held that the contrary view — that the informant cannot be the investigator, and that such overlap entitles the accused to acquittal — is not good law, and it specifically overruled Mohan Lal v. State of Punjab (2018) 17 SCC 627 along with any decision taking the same line. What survives is a fact-sensitive inquiry into actual unfairness, not a categorical bar.

This is a point of some importance for readers navigating the reports, because the name "Mohanlal" appears twice in the NDPS canon in different senses. The case overruled here is Mohan Lal v. State of Punjab (2018), the informant-investigator decision. It is a separate matter from Union of India v. Mohanlal (2016) 3 SCC 379, which laid down directions for the disposal and destruction of seized narcotics under Section 52A of the Act. Mukesh Singh touches only the former; the Section 52A disposal directions remain untouched.

Why the distinction between apprehension and proof matters

The heart of Mukesh Singh is a return to first principles about bias in criminal investigation. Administrative and criminal law have long distinguished between a mere apprehension of bias and demonstrated prejudice. A litigant who says "I fear the process was unfair" states a concern; a litigant who shows that the process was in fact skewed states a case. The Mohan Lal rule collapsed that distinction for one recurring fact pattern, converting a structural feature of narcotics policing into an irrebuttable presumption of unfairness.

The Constitution Bench restored the ordinary discipline. The overlap of informant and investigator is a circumstance the trial court must be alive to — it may call for closer scrutiny of the seizure, the sampling, the independent corroboration and the chain of custody — but it is not, by itself, a verdict. If the accused can show that the officer's dual role in fact tainted the collection or handling of evidence, the trial court can and must act on that. What the accused cannot do is skip the proof and claim acquittal as of right.

This case-by-case framing is deliberately not a blanket endorsement of the practice. The Bench did not say that it is desirable for the informant to investigate his own case, nor that such investigations are beyond challenge. It said only that the challenge must be substantiated. The safeguards that make an NDPS conviction fair — compliance with the search-and-seizure and sampling requirements, and credible corroboration — continue to do the real work; the identity of the officer is one factor within that assessment, not a trump card.

What it means for NDPS trials now

For prosecutors, Mukesh Singh removed a ground on which otherwise sound convictions were being unravelled. A conviction can no longer be set aside on the naked fact that the complainant investigated the case; the defence must point to actual prejudice flowing from that overlap.

For the defence, the decision channels the argument into the evidence. The productive line is no longer "the informant was the investigator, therefore acquit," but "the informant's investigation produced these specific defects" — a broken chain of custody, a compromised sample, an absence of independent witnesses where they were available, non-compliance with the statutory safeguards. Framed that way, the point retains real force; framed as a per se rule, it is now foreclosed.

For trial courts, the judgment sets the method. Treat the overlap as a reason for vigilance, not as a conclusion. Examine whether the investigation was in fact fair, applying the ordinary standards of proof and the NDPS Act's own safeguards, and decide the question of bias or prejudice on the facts of the case before you.

Mukesh Singh thus sits alongside the Court's other NDPS-fairness decisions as a corrective at the level of principle: liberty in narcotics cases is protected by insisting on demonstrated compliance and demonstrated prejudice, not by mechanical presumptions in either direction.

Related on Valkya

• Dr. Rajinder Rajan v. Union of India: grounds-of-arrest discipline reaches NDPS
• Frank Vitus v. Narcotics Control Bureau: bail conditions and the right to privacy
• A.K. Kraipak v. Union of India: the bias rule in natural justice
• Canara Bank v. V.K. Awasthy: prejudice must be shown, not presumed

Sources

1. Supreme Court of India — judgment in Mukesh Singh v. State (Narcotic Branch of Delhi), decided 31 August 2020: api.sci.gov.in judgment PDF
2. Bar & Bench — "NDPS Act: Can informant and investigating officer be the same? Constitution Bench of the Supreme Court answers"
3. LiveLaw — "The Constitution Bench Judgment in Mukesh Singh: A Sensible Retreat from Automatic Bias"
4. S&D Legal Associates — "Case Analysis: Mukesh Singh v. State (Narcotic Branch of Delhi)"