A republic’s travesty of justice
Umar Khalid’s six years in Tihar expose how prolonged pretrial detention under UAPA has become a means of punishing dissent, with Muslims, Dalits and Adivasis bearing a disproportionate burden.

On September 13, 2026, Umar Khalid completed six years in Tihar Jail. Not one prosecution witness has been examined.
There is a particular kind of silence that settles over a case like Umar’s. This is the silence of a system that has learned to punish without verdict, to sentence without trial, and most importantly, to convert the process itself into the penalty. We should not mistake it for the silence of forgetting, as Umar’s name still circulates in seminars, in solidarity statements, and in the anxious conversations of students who understand what his incarceration means for them. Six years. More than two thousand one hundred and ninety days. Not a single witness has been called to testify. And still, no bail.
Miscarriages of justice in the ordinary sense imply error, accident, or the failure of an otherwise functioning mechanism. But what we confront in Umar’s case, and in the broader architecture of which he is the most visible symbol, is something more deliberate. We see the transformation of pretrial detention from an exceptional measure into the standard operating procedure of a government that knows it does not need convictions to put away its dissenters.
Under the Unlawful Activities (Prevention) Act, the conviction rate nationally hovers between 1.5 and 4 per cent. In Jammu and Kashmir, it falls below one per cent. These figures were placed before Parliament and cited in the Supreme Court itself in a ruling this past May, which should be seen as a judicial acknowledgement that one of the country’s most draconian laws produces acquittals in the overwhelming majority of cases. Yet arrests continue to climb. Between 2018 and 2022, over 8,900 UAPA arrests were made, with barely 250 convictions. If the purpose of UAPA was to punish terrorists, it is nothing but a spectacular failure. If its purpose is to punish the accused, to impose suffering before proof, and to discipline dissent through the mere act of arrest, then it is working exactly as designed.
What is it that the draconian UAPA’s Section 43D(5) actually does? Under ordinary criminal procedure, bail is the rule and jail is the exception. But here, the presumption of innocence is inverted, and this presumption is the ancient bulwark of liberty. Under the UAPA, bail cannot be granted if the court is satisfied that a prima facie case exists. And “prima facie” in practice means whatever the prosecution asserts it to mean.
What does all this mean for the question of Umar Khalid specifically? His imprisonment sits at the intersection of two threat categories the current dispensation treats interchangeably. First is being Muslim, and second is being a well-educated, articulate critic of majoritarian state power.
The prosecution’s submissions characterised the 2020 CAA-NRC protests as an “orchestrated” attack on India’s sovereignty rather than a spontaneous civic response to an arbitrary and exclusionary law. This rhetorical move was aimed at criminalising dissent itself as sedition-adjacent conduct. Five co-accused in the same case, Gulfisha Fatima, Meeran Haider, Shifa-ur-Rehman, Saleem Khan, and Shadab Ahmed, were granted bail in January 2026. Khalid Saifi and Tasleem Ahmed walked out in May 2026. However, Umar Khalid and Sharjeel Imam alone were held on a “qualitatively different footing".
Politically, this phrase manufactures a hierarchy of suspicion mapped almost perfectly onto who is seen as the “ideological” Muslim defendant versus the merely “implicated” one. Justice Deepak Gupta, in his recent article, sheds light on something similar where members of one community were denied bail for months for eating chicken on a boat on the Ganga, while members of the majority community drinking liquor on the same river were bailed within hours. The visceral, concrete absurdity of this juxtaposition captures something essential about how the law actually operates.
In November 2012, a CPI(M) delegation led by general secretary Prakash Karat, along with Subhashini Ali and Mohd. Yusuf Tarigami, met President Pranab Mukherjee over the mass wrongful implication of Muslim youth in terror cases. Mohammed Aamir Khan of Delhi, jailed 14 years and acquitted in 17 cases. Syed Maqbool Shah of Srinagar, 14 years. Syed Wasif Haider of Kanpur, 8 years. All were eventually vindicated by courts that found the “terrorist” narrative fabricated.
This history is important because it establishes that the pattern of anti-Muslim targeting predates the current government. However, the current BJP-RSS regime brought a change in the scale, intensity, impunity, and ideological normalisation of this pattern.
And yet, in Tihar, Umar’s life has been reduced to small consolations like his companionship with two prison cats, newspapers, and weekly calls, which are his only lifeline to hope. This serves as evidence of what “process as punishment” costs in lived, daily terms. If the process itself consumes six, eight, or ten years and ends overwhelmingly in acquittal, the trial is effectively replaced by the punishment.
The Bhima-Koregaon case offers a parallel indictment of the same machinery. It has been eight years since the alleged incident. Even the trial has not begun. One accused, Father Stan Swamy, an 84-year-old tribal rights activist with Parkinson’s disease, died in custody on July 5, 2021, a day before his bail hearing. He was denied even a straw and a sipper cup to drink water. His death was an institutional killing through a calculated denial of care.
Anand Teltumbde, the Dalit scholar and IIT professor, was denied bail days after Swamy’s death. When the Bombay High Court finally granted him bail in November 2022, it found the NIA’s material against him to be in the “realm of speculation”. Let that phrase sit for a moment! His years of incarceration were based solely on speculation, if not on a conspiracy against his dissent with the ruling party. Six senior citizens, including Teltumbde, were held for years pending trial, several with serious comorbidities.
The idea is to arrest the annotators of caste and capital to send a message to every university department and NGO in the country, without a single conviction even being necessary. The organic intellectuals have been in the crosshairs of the RSS-led regime because they translate subaltern grievance into organised political language. Imprisoning Teltumbde was an attack on the Ambedkarite-Marxist intellectual tradition itself.
For a ‘naxal-free India’
But this story is not limited to the English-speaking, urban political prisoners. The statistics from Bastar are even starker. Over 180 Adivasis were killed in “encounters” in 2024 alone, by one estimate. With the state setting a formal deadline to establish a “naxal-free India by March 2026”, the body count rose past 450 by mid-2025, according to international rights organisations. What does a body-count deadline do to due process? What does the presumption of innocence mean for anyone who happens to live in a conflict zone?
Take the case of Suneeta Pottam, a 25-year-old Adivasi rights defender who organised protests against illegal security camps and documented extrajudicial killings and sexual violence. She was arrested and branded a Maoist under the UAPA, just a day after the 2024 Lok Sabha election results. Or consider the Moolvasi Bachao Manch arrests, which included Sarju Tekam and Raghu Midiyami, among others who were jailed under the UAPA and CSPSA for organising against illegal mining, militarisation, and forced land acquisition. It is not difficult to notice that the label “Maoist” functions in Bastar exactly as “urban naxal” functions in Bhima Koregaon and “anti-national” functions in Delhi. It is a linguistic device that preempts the need for evidence by pre-loading the accusation with terror.
The so-called security rhetoric is a cover-up for the material reality where the proliferation of security camps in Fifth Schedule areas tracks the geography of iron ore and bauxite reserves. The Union Environment Minister Bhupender Yadav himself stated in Parliament that between 2014-15 and 2023-24, a total of 1,73,396.87 hectares of forest land had been diverted for non-forest use. The neo-fascist regime of the BJP-RSS is deploying the resources of the state to facilitate the plunder of natural resources for its corporate friends, a 21st-century scheme of primitive accumulation enforced by the guns of the state.
The National Crime Records Bureau (NCRB) data provides a mountain of evidence for this indictment. Prison Statistics India 2023 shows that out of the 5.8 lakh total prisoners, 77.9 per cent (approximately 4.5 lakh) are undertrials, the highest proportion in over a decade. The national occupancy is at 133 per cent of capacity. Dalits constitute 22 per cent of all prisoners against roughly 16.6 per cent of the population. Adivasis 13 per cent against 8.6 per cent. Muslims 16 per cent against roughly 14.2 per cent. These disproportions have barely moved across successive NCRB reports since 2015, when the combined Muslim-Dalit-Adivasi share of undertrials crossed 55 per cent against a combined 39 per cent population share.
If a prison system reflects India’s caste and religious hierarchy this accurately, then it should be seen as a mirror held up to the Manuvaadi social order that the ruling bloc wishes to preserve. And we should note the 2016-2017 data gap where the NCRB inexplicably dropped caste and religion breakdowns for those two years. The NCRB also does not record cases of mob lynchings anymore.
On the one hand the state purposefully denies justice by keeping innocent people behind bars, on the other hand we see convicted rapists like Ram Rahim, Asaram, and the rapists of Bilkis Bano, who are either released early or on parole because of their political proximity to the ones in power. They do not even have to serve their full sentence and receive a hero’s welcome every time they step out of jail.
This brings us to the question of Manuvaad or the caste hierarchy as social ordering. The RSS-BJP combine does not need to legislate the Manusmriti into formal law. It achieves the same result informally, through selective policing, selective bail, selective encounter-killing, and selective media vilification. A soft, deniable Manuvaad enacted through the administrative and carceral state is a much better tool than the statute book, especially when the optics are concerned.
If caste and religious hierarchy have no legal standing in the Republic of India, why does its prison population reproduce that exact hierarchy with such fidelity? And the bench itself reflects this ordering. Recent Supreme Court and High Court Chief Justice appointments have skewed heavily upper caste despite no shortage of qualified judges from other communities among the top fifty High Court judges in seniority.
Consider, too, the tale of two judiciaries. Economic offenders with resources to litigate for years obtain anticipatory bail, settle via compounding, or simply leave the country. UAPA undertrials are denied bail for six-plus years on unproven allegations. The top 10 per cent of Indians own about 65 per cent of national wealth and 57.7 per cent of national income; the bottom 50 per cent own just 6.4 per cent of the national wealth and 15 per cent of the national income. Similarly, there is a significant under-representation of these communities when landownership is concerned. Unequal access to liberty before the law is simply the judicial expression of unequal access to capital. The judiciary, as part of the superstructure, reproduces the class relations of the base. “Equality before law” is formally guaranteed and substantively negated by capital’s capacity to purchase delay, expertise, and reasonable doubt.
Giorgio Agamben identified this as the state of exception becoming the rule. The emergency architecture becomes the ordinary mode of governance for targeted populations. For Muslims in Delhi, for Adivasi activists in Bastar, and for Dalit scholars in Maharashtra, the exception is the everyday, where instead of protecting them as law-abiding citizens, the law processes them as criminals.
The Marxist-Leninist understanding of the state as a class instrument ties these disparate threads together. The judicial and police apparatus, far from being a neutral arbiter, serves the reproduction of ruling-class and dominant-caste power. The UAPA serves as the sharpest current edge of this instrument. Gramsci’s concept of hegemony explains how the RSS’ decades-long project of civil society and state apparatus penetration (education, media, judiciary appointments) manufactures consent for the coercive apparatus to act without triggering broad public resistance. Imprisoning Umar becomes possible because a hegemonic ‘common sense’ has already framed dissent as anti-national. Similarly, Ambedkar’s warning on constitutional morality tells us that a Constitution’s text guarantees nothing if the underlying society remains saturated with graded inequality. Caste and class are intertwined axes of domination.
The resistance from Communists, Muslims, Dalits, and Adivasis forms the front whose historical dissent most directly threatens the Manuvaadi-corporate consensus the RSS project seeks to entrench. In Delhi, the UAPA targets Muslim protesters against the CAA-NRC. In Bhima Koregaon, it criminalises Ambedkarite intellectuals. In Bastar, it labels indigenous self-organisation against corporate plunder as Naxalism or Maoism.
The 2012 CPI(M) memorandum’s four demands still form the policy spine that we must revive. The demands were: (a) compensation and rehabilitation for the wrongfully implicated; (b) special fast-track courts with time-bound closure within a year; (c) accountability through prosecution of investigators found to have concocted evidence; and (d) reconsideration and repeal of UAPA’s draconian provisions. Not one of these demands has been implemented in the fourteen years since. If anything, UAPA’s punitive architecture has only been strengthened. The 2019 amendments allow individuals, not just organisations, to be designated as “terrorists”.
So what is to be done? We must revive, update, and reissue the four demands as 2026 demands: compensation and rehabilitation; time-bound special courts; accountability for fabricated evidence, naming this explicitly against current investigating agencies; and repealing of draconian laws like the UAPA and Section 152 of the Bharatiya Nyaya Sanhita, which is worse than the old sedition law. Our new demands should also include an independent audit of NCRB caste and religion prison data with mandatory annual disclosure; judicial appointment reform addressing the representational imbalance; and immediate release of ailing and elderly undertrials on humanitarian bail as a floor, not an exception won case by case.
Six years of Umar Khalid’s life cannot be restored by an eventual acquittal. The same is true for the Bhima Koregaon accused, for Suneeta Pottam, and for thousands of nameless Muslim, Dalit, and Adivasi undertrials and other political prisoners. Time in a cell is not compensable. This is why the process-as-punishment doctrine must be named and fought as one of the most prominent human rights crises of this period.
We must anticipate that the state will attempt to counter by invoking national security, prima facie findings, and sub judice protocols. But we must also remember that this is a political battle and should be fought with sustained political pressure and, not with the hope of judicial goodwill or token judicial reforms. The left and democratic forces of the country are committed to steering this battle towards victory. The Indian freedom fighters did not give their lives for a republic that jails first and asks questions later, nor was this republic promised to us by our Constitution.
Six years is long enough to know the difference.
M.A. Baby is the general secretary of Communist Party of India (Marxist).