Ashwini Kumar Upadhyay v. Union of India: dismissing the hate-speech batch
On 29 April 2026, a two-judge bench dismissed thirteen writs, two SLPs and eight contempts in the long-running hate-speech batch, holding that constitutional courts cannot create criminal offences, that no legislative vacuum exists in the IPC/BNS framework, and that police failure to register a suo motu FIR is not, by itself, contempt.
- Court
- Supreme Court of India
- Citation
- 2026 INSC 432
- Bench
- Vikram Nath, J., Sandeep Mehta, J.
- Decided
- 29 April 2026
The facts in brief
From 2020 onwards the Supreme Court had been seized of a clutch of writ petitions filed under Article 32 by Ashwini Kumar Upadhyay — a Delhi-based advocate and serial PIL litigant — by Shaheen Abdullah, by the All India Majlis-e-Ittehadul Muslimeen and by other petitioners seeking wide-ranging directions on hate speech and communal incitement. The petitions multiplied through 2021, 2022 and 2023 as fresh incidents were reported across multiple states. By early 2026 the batch had grown to thirteen writ petitions, two SLPs (one civil, one criminal) and eight contempt petitions — the contempts arising from alleged non-compliance with interim directions issued in Shaheen Abdullah v. Union of India in 2022.
In Shaheen Abdullah, a bench then headed by Justice K.M. Joseph had directed police authorities to register suo motu FIRs in hate-speech cases without waiting for a complaint, and had warned of contempt for non-compliance. As compliance proved patchy across states, contempt petitions were filed. The Vikram Nath / Sandeep Mehta bench took up the consolidated batch and heard extended arguments on the petitioners' substantive prayers — mandatory suo motu FIR registration, court-monitored special-investigation-team supervision, judicial implementation of the Law Commission of India's 267th Report on Hate Speech (March 2017), and continuing mandamus on the executive's enforcement obligations.
Judgment was reserved in March 2026. On 29 April 2026 the bench delivered a single comprehensive judgment, authored by Vikram Nath J., dismissing the entire batch save for one Noida-assault matter that was kept pending for separate disposal on its individual facts.
The constitutional question
Three questions were doctrinally tied together. First, whether the constitutional court may direct the creation or expansion of criminal offences — through judicial declaration, guidelines or continuing mandamus — where Parliament has not legislated. Second, whether the existing IPC framework on hate speech (and its successor in the Bharatiya Nyaya Sanhita 2023) suffers a "legislative vacuum" that warrants extraordinary judicial intervention. Third, whether police failure to register a suo motu FIR in hate-speech cases is, by itself, contempt of the Shaheen Abdullah (2022) directions.
The three questions sit on a single axis: how far constitutional courts may travel in disciplining an executive function that Parliament has chosen to structure through ordinary penal legislation.
What the Court held
Separation of powers in the criminal-law domain
The bench held that the judiciary cannot assume a law-making function in the criminal-law domain. Article 20(1)'s bar on ex post facto criminalisation operates as an additional structural constraint: a person cannot be punished for an offence that was not on the statute book at the time of the conduct, and a court that creates or expands a criminal offence by judicial declaration would inevitably operate retrospectively in respect of conduct already engaged in. The structural objection is therefore not merely separation-of-powers in the abstract; it tracks a specific constitutional limit on the criminal law's reach.
A constitutional court cannot create or expand criminal offences in the absence of legislative sanction.
The statement is short but doctrinally consequential. It does not merely decline a particular prayer — it articulates a general limit on the kind of relief that PIL petitioners may seek in the criminal-law domain.
"No legislative vacuum"
The bench surveyed the existing IPC framework — sections 153A (promoting enmity between groups), 153B (imputations prejudicial to national integration), 295A (deliberate and malicious acts intended to outrage religious feelings), 298 (uttering words with deliberate intent to wound religious feelings) and 505 (statements conducing to public mischief) — and observed that those provisions had been carried forward into the Bharatiya Nyaya Sanhita 2023 as sections 196, 197, 299 and 351 with only minor architectural adjustments.
The existing framework sufficiently covers provisions that deal with speeches that outrage religious sentiments or disturb public tranquillity; the issue is one of enforcement, not a legislative vacuum.
The doctrinal move is the recharacterisation. The petitioners had framed the hate-speech problem as one of doctrinal inadequacy — the existing provisions were said to be inadequate, badly drafted or insufficiently used. The bench rebooted the diagnosis: the provisions exist, the legislative architecture is in place, the deficiency is administrative. Police training, prosecutorial discipline, internal-affairs accountability — these are the loci of reform, not new judicial offences.
The Law Commission's 267th Report
The petitioners had urged judicial implementation of the Law Commission of India's 267th Report on Hate Speech, which had been submitted in March 2017 and which had recommended — among other things — fresh penal provisions specifically targeting incitement to hatred. The bench declined. The Report's recommendations, however well-considered, remain in the legislative domain; they are commendable inputs for Parliament's consideration but cannot be translated into directly-enforceable judicial directions. The doctrinal point is the same as the "no legislative vacuum" point in different clothing — the Court will not act as a surrogate legislature, even where the surrogate inputs are well-prepared.
Police failure to register a suo motu FIR is not automatically contempt
The most operationally significant holding addresses the contempt jurisdiction. The bench held that the failure of police to suo motu register an FIR for hate speech is not, by itself, automatically contempt of court. The contempt jurisdiction requires a specific complaint that has gone uninvestigated, not generalised non-compliance with a broad suo motu duty.
The move doctrinally moderates the Shaheen Abdullah (2022) interim direction. Shaheen Abdullah had imposed a wide suo motu FIR obligation and had threatened contempt for non-compliance. The Vikram Nath / Sandeep Mehta bench did not overrule that direction. It disciplined the contempt consequence: the broad suo motu obligation may continue to exist as administrative discipline, but a generalised failure to register a suo motu FIR will not, without more, trigger contempt. A specific-complaint-not-acted-upon predicate is required.
Disposal of the batch
The bench dismissed the thirteen writ petitions and the two SLPs. It closed the eight contempt petitions. It kept pending the Noida assault matter — a specific incident involving an assault on a Muslim man in Noida, Uttar Pradesh — for separate disposal on its individual facts, recognising that the Noida matter could be assessed within the specific-complaint framework that the contempt holding had clarified. The bench did not foreclose the legislative path: Parliament remains free to consider the Law Commission's recommendations and to legislate fresh provisions if it so chooses.
The doctrinal architecture
The judgment accomplishes three doctrinal moves in a single text.
First, it restates the separation-of-powers limit in the criminal-law domain with unusual clarity. Earlier judgments — including Subramanian Swamy v. Union of India (2016) on criminal defamation and Shreya Singhal v. Union of India (2015) on Section 66A of the IT Act — had walked the line on judicial supervision of speech-criminalisation. Ashwini Upadhyay (2026) supplies the explicit doctrinal statement: courts may strike down offences that fail constitutional scrutiny, but they may not create or expand offences to fill perceived gaps.
Second, it articulates the "no legislative vacuum" doctrine as a gating principle for PIL relief. Future Article 32 petitioners who seek judicial creation of criminal offences will need to demonstrate, in the first instance, that the existing penal architecture is doctrinally inadequate — not merely under-enforced. The diagnosis is administrative, not architectural, in the bench's view; PIL relief calibrated to administrative deficits looks very different from PIL relief calibrated to architectural deficits.
Third, it disciplines the contempt jurisdiction in continuing-mandamus PIL contexts. Where the Court has issued broad directions on executive action — and where compliance has proved generalised rather than specific — the contempt sanction will not automatically attach. The petitioner who seeks contempt must identify a specific complaint that has gone uninvestigated, not merely a generalised pattern of inaction.
What the judgment did not decide
The judgment did not address the constitutionality of any specific hate-speech provision. The IPC and BNS frameworks were treated as constitutionally adequate; the bench was not invited to, and did not, examine whether sections 196, 197, 299 or 351 BNS suffer Article 19(1)(a) defects. That question awaits a future as-applied or facial challenge.
The judgment did not foreclose the Shaheen Abdullah (2022) interim directions on suo motu FIR registration. The directions remain on the books as administrative discipline; what changed is the contempt consequence, not the underlying obligation. Police authorities may continue to be expected to register suo motu FIRs where the facts warrant — but failure to do so will not, without a specific-complaint predicate, attract contempt.
The judgment did not resolve the Noida assault matter on its facts. That matter was kept pending and will be disposed of separately. The disposal of that single matter will not disturb the doctrinal framework that the present judgment has now laid down.
After the judgment
The judgment has begun to reshape Article 32 PIL practice in the speech-regulation space. Future petitioners are recalibrating their prayers toward enforcement-focused relief — specific-complaint-inaction directions, police-accountability mechanisms, prosecutorial-discipline frameworks — rather than toward judge-made-offence-architecture prayers. The shift is significant: a decade of hate-speech PIL had been organised around the assumption that the existing penal framework was inadequate and that judicial supplementation was both possible and necessary. Ashwini Upadhyay (2026) closes that route.
The judgment has attracted civil-liberties criticism. The Shaheen Abdullah (2022) suo motu FIR directions had been welcomed by some commentators as a structural intervention against police inaction in communal-incitement cases; the doctrinal moderation of the contempt consequence is read by those commentators as a step backward. The bench's framing — enforcement-not-vacuum — is contested in the academic literature.
Parliamentary attention may turn to whether the Law Commission's 267th Report should be legislatively taken up. The Standing Committee on Personnel, Public Grievances, Law and Justice has signalled interest. Whether the political appetite exists for fresh hate-speech legislation in the present environment is an open question; the judgment squarely places that question in the legislative domain.
The separation-of-powers reasoning is also being cited in PIL contexts well beyond hate speech — in proposed offences in cyber-harassment, environmental crimes and gender-violence contexts, the Ashwini Upadhyay discipline is now the gating principle that PIL petitioners must navigate.
Related on Valkya
- Shreya Singhal v. Union of India: striking down Section 66A
- K.S. Puttaswamy v. Union of India: the nine-judge privacy declaration
- Anuradha Bhasin v. Union of India: internet as an Article 19 right
Sources
- Verdictum — Ashwini Kumar Upadhyay v. Union of India case-page (2026 INSC 432): https://www.verdictum.in/supreme-court/declines-directions-on-hate-speech-prior-sanction-not-firs-ashwani-upadhyay-1613032
- LiveLaw — 2026 LiveLaw (SC) 437 report: https://www.livelaw.in/amp/sc-judgments/2026-livelaw-sc-437-ashwini-kumar-upadhyaya-v-union-of-india-532271
- LiveLaw — non-registration of suo motu FIR no contempt analysis: https://www.livelaw.in/top-stories/supreme-court-hate-speech-batch-non-registration-of-suo-motu-fir-no-contempt-if-complaint-not-made-to-police-failure-to-act-necessary-532917
- Supreme Court Observer — SCO.LR Vol 5 Issue 1 reference page: https://www.scobserver.in/
- Law Commission of India — Report No. 267 on Hate Speech (March 2017): https://lawcommissionofindia.nic.in/
Related reading
Shreya Singhal v. Union of India: striking down Section 66A
Navtej Singh Johar v. Union of India: the unanimous decriminalisation of consensual same-sex conduct
Joseph Shine v. Union of India: the unanimous striking down of Section 497 IPC
Trace how this proposition has been treated across Indian courts — citations, bench strength, and subsequent history — in one workspace built for litigators.