Legal Architecture of Internal Security

The constitutional framework established where the power comes from. This section examines what Parliament and the State legislatures have actually built on that foundation. It is the longest section of the chapter, and the most heavily examined, because every one of these statutes sits on the fault line identified at the outset: security and liberty are conditions of each other, and a law that purchases one by destroying the other has failed even on its own terms.

Read the statutes as a family, not as isolated Acts. Each was enacted in response to a specific failure, each departs from ordinary criminal procedure at a specific point, and each departure — longer detention before charge, reversed presumptions, restricted bail, immunity from prosecution — is the precise place where the constitutional argument happens. Learn the departure, and you can argue any question on any of them.

The Unlawful Activities (Prevention) Act, 1967

Origin and evolution

The UAPA was enacted in 1967, following the Sixteenth Amendment, 1963, which added the sovereignty and integrity of India as a ground for restricting the freedoms under Article 19. Its original target was secessionist activity — “unlawful activity” meaning action supporting cession or secession, or questioning or disrupting the sovereignty and territorial integrity of India. Terrorism was not originally part of it.

The Act became India’s principal anti‑terror statute only by absorption, after two dedicated statutes were enacted and then abandoned:

UAPA 1967

→

TADA 1985/87 lapsed 1995

→

POTA 2002 repealed 2004

→

UAPA absorbs terror

TADA lapsed in 1995 amid findings of mass arrests with negligible convictions; POTA was repealed in 2004 on similar grounds. On each occasion the substantive content migrated into the UAPA, which — being an ordinary statute with no sunset clause — has never had to justify its own renewal. This is a point worth making: India did not choose the UAPA as its anti‑terror law; it arrived there by default, after the laws that were subject to review were allowed to die. Major amendments followed in 2004, 2008 (after 26/11), 2012 and 2019.

Key provisions

Provision	Content
Unlawful activity	Action supporting cession or secession, disclaiming or questioning the sovereignty and territorial integrity of India, or causing disaffection against India.
Terrorist act (S.15)	Acts done with intent to threaten or likely to threaten the unity, integrity, security, economic security or sovereignty of India, or to strike terror in the people, using bombs, firearms, hazardous substances or other means.
Ban on organisations (S.3, First Schedule)	The Central Government may declare an organisation unlawful, and may list terrorist organisations in the First Schedule.
Designation of individuals (S.35, Fourth Schedule)	Inserted by the 2019 amendment. The Central Government may notify an individual as a terrorist and list him in the Fourth Schedule.
Attachment of proceeds (S.25)	Property representing proceeds of terrorism may be seized or attached. Post‑2019, where investigation is by the NIA, approval of the Director General, NIA suffices.
Investigation (S.43)	Post‑2019, officers of the rank of Inspector and above in the NIA may investigate, in addition to the previously prescribed higher ranks.
Extended custody and charge‑sheet (S.43D)	Time for filing the charge‑sheet may be extended to 180 days; police custody may extend beyond the ordinary limit.
Bail restriction (S.43D(5))	Bail shall not be granted if the court, on a perusal of the case diary or charge‑sheet, is of opinion that there are reasonable grounds for believing the accusation is prima facie true.

The 2019 amendment — four changes

• Designation of individuals as terrorists under Section 35, listed in the Fourth Schedule. Previously only organisations could be proscribed.
• Property attachment on the approval of the Director General of the NIA, removing the earlier requirement of prior permission from the State police chief where the NIA is investigating.
• Investigation by NIA officers of Inspector rank and above under Section 43.
• Addition of the International Convention for the Suppression of Acts of Nuclear Terrorism to the Schedule of treaties.

THE ARGUMENT ON INDIVIDUAL DESIGNATION The case for it. Terrorist organisations dissolve and reconstitute under new names the moment they are proscribed; individuals do not. Designating the individual attaches consequences — asset freezing, travel restriction, a basis for extradition requests and for international listing — to the person who actually organises violence, rather than to a label he can discard. It also aligns Indian practice with United Nations sanctions listing, which has always been individual as well as organisational.   The case against it. Designation is an executive act, made without a hearing, without a judicial finding, and without any requirement of conviction or even of charge. The only remedy is an application to the Central Government for de‑notification, and thereafter to a review committee constituted by the Central Government itself — the same executive that made the designation. Critics argue this offends the principle that a person may not be branded a criminal without adjudication, and note that the stigma and consequences attach immediately and are practically irreversible.   The status. Constitutional challenges to the 2019 amendment are pending before the Supreme Court.

The evidence problem

Any serious answer on the UAPA must engage the data, because it is the strongest available criticism and it comes from the government’s own returns:

• Between 2019 and 2023, approximately 10,440 persons were arrested under the UAPA, and approximately 335 were convicted — a conviction rate in the range of 1.5 to 4 per cent.
• Pendency is the mechanism of the harm. With investigation pendency around 81 per cent and trial pendency around 89 per cent, and with Section 43D(5) making bail exceptionally difficult, the process becomes the punishment: a person may spend years in custody and be acquitted.
• The bail standard was set in NIA v. Zahoor Ahmad Shah Watali (2019), where the Supreme Court held that at the bail stage a court must accept the material in the charge‑sheet broadly at face value and must not conduct a detailed examination of its probative value — which, in practice, makes the prosecution’s own narrative close to conclusive.
• The counterweight is Union of India v. K. A. Najeeb (2021), where the Court held that constitutional courts may grant bail notwithstanding Section 43D(5) where prolonged incarceration without prospect of an early trial violates Article 21. The Supreme Court has since reaffirmed this line and cautioned against readings of Watali that hollow out constitutional protection.

The balanced position. The State’s case is real: terrorism cases depend on intelligence that is often not producible as evidence, on witnesses who will not testify, and on conspiracies that must be disrupted before completion — so a low conviction rate is not by itself proof of misuse.

But a 1.5 to 4 per cent conviction rate sustained over five years, combined with multi‑year pre‑trial detention, is not a marginal imperfection; it is evidence that the statute is functioning as a detention mechanism rather than a prosecution mechanism. Both propositions can be stated in the same answer, and the better answer states both.

The National Investigation Agency Act, 2008

Enacted in the weeks after the 26/11 Mumbai attacks, the NIA Act created India’s first genuinely federal investigative agency for terrorism. Its jurisdiction is defined not by territory but by a Schedule of offences — principally under the UAPA, the Atomic Energy Act, the Anti‑Hijacking Act, the SAARC Convention (Suppression of Terrorism) Act and related statutes.

Where a scheduled offence is registered, the Central Government may direct the NIA to take over the investigation, and the State police must transfer the case.

The 2019 amendment

Change	Content
Extraterritorial jurisdiction (S.1(2))	The NIA may investigate scheduled offences committed outside India — by Indian citizens, against Indian citizens, or affecting India’s interests — and in aircraft or ships registered in India, as if the offence had been committed in India, subject to international treaties and the domestic law of the country concerned.
New scheduled offences	Added human trafficking, counterfeit currency and banknotes, manufacture or sale of prohibited arms, cyber terrorism (Section 66F of the Information Technology Act, 2000) and offences under the Explosive Substances Act, 1908.
Special Courts	Empowered the Central Government, and also State Governments, to designate Sessions Courts as Special Courts for the trial of scheduled offences, addressing the earlier bottleneck of too few designated courts.

The federal objection, which belongs in any answer on this Act: the NIA may assume an investigation without the concurrence of the State, displacing the State police from a case arising on their own territory, on a subject — police — that is constitutionally a State entry. The Union’s answer is that terrorism is by nature inter‑State and transnational and that an agency requiring thirty‑odd consents would be useless. Both propositions are sound; the absence of any consultative mechanism between them is the defect.

The Armed Forces (Special Powers) Act, 1958

Act No. 28 of 1958, dated 11 September 1958, originally titled the Armed Forces (Assam and Manipur) Special Powers Act, 1958 and renamed in 1972. It extends to Arunachal Pradesh, Assam, Manipur, Meghalaya, Mizoram, Nagaland and Tripura. A separate statute, the Armed Forces (Jammu and Kashmir) Special Powers Act, 1990, applies there. The Act is short — six operative sections — and every one of them matters.

The operative provisions

Section	Content
S.2(b)	Defines “disturbed area” as an area declared as such by notification under Section 3.
S.3	Power to declare. Where the Governor of a State, the Administrator of a Union territory, or the Central Government is of opinion that the whole or any part of the State or Union territory is in such a disturbed or dangerous condition that the use of armed forces in aid of the civil power is necessary, that authority may by notification in the Official Gazette declare the whole or any part of it a disturbed area.
S.4(a)	Any commissioned, warrant or non‑commissioned officer or person of equivalent rank may, after giving such due warning as he may consider necessary, fire upon or otherwise use force, even to the causing of death, against any person acting in contravention of any law or order for the time being in force in the disturbed area prohibiting the assembly of five or more persons, or the carrying of weapons, firearms, ammunition or explosive substances — if he is of opinion that it is necessary to do so for the maintenance of public order.
S.4(b)	Destroy any arms dump, prepared or fortified position or shelter from which armed attacks are made or are likely to be made, or any structure used as a training camp for armed volunteers or as a hide‑out by armed gangs or absconders.
S.4(c)	Arrest without warrant any person who has committed a cognizable offence, or against whom a reasonable suspicion exists that he has committed or is about to commit a cognizable offence, using such force as may be necessary.
S.4(d)	Enter and search without warrant any premises to make such an arrest, to recover a person wrongfully confined, or to recover stolen property, arms, ammunition or explosive substances.
S.5	Any person arrested must be made over to the officer in charge of the nearest police station with the least possible delay, together with a report of the circumstances of the arrest.
S.6	Protection. No prosecution, suit or other legal proceeding shall be instituted, except with the previous sanction of the Central Government, against any person in respect of anything done or purported to be done in exercise of the powers conferred by the Act.

WHERE THE OBJECTION ACTUALLY LIES The activists’ case is not that soldiers should be defenceless. It is directed at four specific features, and an answer that names them precisely will always outperform one that calls the Act “draconian” in general terms: Section 4(a) permits the causing of death on the subjective opinion of a non‑commissioned officer, for the maintenance of public order, against a person contravening an order prohibiting assembly of five or more — a threshold far below the imminent‑threat standard that governs the use of lethal force elsewhere. Section 4(c) permits arrest on suspicion that a person “is about to commit” an offence — a pre‑emptive standard with no external check. Section 6 makes prosecution conditional on the Central Government’s sanction, which in practice is rarely granted, so that accountability depends on the discretion of the same executive that ordered the deployment. Section 3 attaches no time limit and no independent review to the declaration of a disturbed area, which is why notifications made for an emergency have been renewed for decades.

The judicial position

• Naga People’s Movement of Human Rights v. Union of India (1997). A Constitution Bench upheld the constitutionality of the Act, but subjected it to important conditions: the power under Section 3 is not arbitrary and the declaration must be for a limited duration with periodic review; the “due warning” under Section 4(a) is mandatory; only the minimum force necessary may be used; the arrested person must be handed to the police with least possible delay; and the Army’s own “Dos and Don’ts” are binding and their violation invites action under the Army Act.
• Extra Judicial Execution Victim Families Association (EEVFAM) v. Union of India (2016). The Court held that the use of excessive force is impermissible even in a disturbed area and even against militants, insurgents and terrorists; that Section 6 confers no absolute immunity, since immunity cannot be determined before it is established whether an encounter was genuine; and that every death caused by the armed forces must be thoroughly enquired into. This is the single most important AFSPA authority for a Mains answer, because it converts the accountability argument from a political demand into a judicial holding.

The committee record

Committee	Recommendation
Justice B. P. Jeevan Reddy Committee, 2005	Recommended that AFSPA be repealed forthwith, describing it as “a symbol of oppression, an object of hate and an instrument of discrimination and high‑handedness”. It did not recommend disarming the State: it proposed that the necessary powers be incorporated, with safeguards, into the Unlawful Activities (Prevention) Act, 1967.
Justice Santosh Hegde Committee, 2013	Constituted on the Supreme Court’s direction to examine encounter deaths in Manipur. Found that five of the six encounters examined were not genuine, that disproportionate force had been used against persons with no known criminal antecedents, and that the Act confers sweeping powers without adequate protection against misuse.
Second Administrative Reforms Commission	Also recommended repeal of the Act, with the necessary provisions to be relocated into ordinary security legislation.

The counter‑argument deserves fair statement. The armed forces’ position is that soldiers deployed in counter‑insurgency are performing a task for which they were not trained and are not legally equipped as police officers; that without Section 6 every operational decision would be litigated by hostile parties, producing paralysis in the field; and that the Act has never been the cause of the underlying conflict.

On this view, the demand should be for better operating procedures and faster sanction decisions, not for repeal. A balanced answer records this and then observes that the two positions are not irreconcilable: Jeevan Reddy’s proposal — relocate the powers into a statute with built‑in safeguards, time limits and accountability — grants the operational necessity while removing the impunity.

The rollback

The most significant development of the last decade is the progressive contraction of disturbed‑area notifications — a fact that changes the shape of the AFSPA answer entirely, because the debate is no longer purely abstract.

• Tripura withdrew the notification in 2015; Meghalaya in 2018; Mizoram earlier.
• In Assam, the disturbed‑area declaration has been withdrawn from 23 districts, and now applies to nine districts — Tinsukia, Dibrugarh, Charaideo, Sivasagar, Jorhat, Golaghat, Karbi Anglong, West Karbi Anglong and Dima Hasao — and the Lakhipur sub‑division of Cachar.
• Manipur remains a disturbed area except for the jurisdictions of thirteen police stations across five districts.
• Nagaland and Arunachal Pradesh retain notifications over specified districts and police‑station areas, extended in six‑month cycles.
• Taken together, AFSPA now applies fully in about 31 districts and partially in about 12 districts across four North‑Eastern States, against near‑total coverage two decades ago.

In the vocabulary of securitisation theory, the rollback is a deliberate act of desecuritisation — returning territory from exceptional to ordinary law as the security situation permits. It is the strongest available evidence that the Indian state treats these powers as contingent rather than permanent, and it is the point to make when a question asks whether AFSPA can ever be withdrawn.

The National Security Act, 1980, and Preventive Detention

The NSA is the principal central preventive detention statute, enacted under Entry 9 of the Union List and Entry 3 of the Concurrent List. Preventive detention is not punishment for an offence committed; it is restraint against an apprehended future act, and therefore operates entirely outside the criminal trial.

Feature	Content
Grounds	Detention to prevent a person from acting in a manner prejudicial to the defence of India, the relations of India with foreign powers, the security of India; or the security of the State, the maintenance of public order, or the maintenance of supplies and services essential to the community.
Detaining authority	The Central Government, a State Government, or — where empowered — a District Magistrate or Commissioner of Police.
Maximum period	Twelve months from the date of detention, subject to the order being revised or revoked earlier.
Advisory Board	Constituted of three persons who are, have been, or are qualified to be Judges of a High Court. The detention must be referred to the Board within three weeks, and the Board reports whether there is sufficient cause for detention.
Grounds of detention	Must be communicated to the detenu and he must be afforded the earliest opportunity of making a representation against the order.
Legal representation	The detenu has no right to be represented by a legal practitioner in proceedings before the Advisory Board.

Jurisprudential position. The courts have not struck preventive detention down — Article 22 authorises it — but have policed its edges strictly, insisting that the grounds be communicated promptly and intelligibly, that they not be vague, irrelevant or stale, that the representation be considered by the detaining authority without avoidable delay, and, critically, that “public order” be distinguished from “law and order”: an act that disturbs the ordinary tranquillity of a locality is not automatically a threat to public order, and detention on that basis will be quashed. Where a person is already in custody, a fresh detention order requires the authority to record awareness of that fact and a reasoned satisfaction that he is likely to be released and to act prejudicially.

The structural criticism is nevertheless powerful. The detenu is deprived of liberty for up to a year on the executive’s satisfaction, reviewed by a body that is advisory and not judicial, before which he cannot be represented by counsel, on grounds from which facts against the public interest may be withheld under Article 22(6).

Recall that the safeguard Parliament enacted in the Forty‑fourth Amendment — reducing the pre‑Board period from three months to two — has never been brought into force. The State’s answer is that an organised network cannot be disrupted by prosecution alone and that detention is preventive rather than punitive; the answer to the answer is that a preventive power without adjudication is, in effect, punishment without trial.

State Legislation

Because criminal law and preventive detention for public order both sit in the Concurrent List, States have built their own security statutes. Where these depart from central law, they require the assent of the President under Article 254(2).

Statute	Reach	Distinctive features
Maharashtra Control of Organised Crime Act, 1999 (MCOCA)	Maharashtra; extended to Delhi.	Targets the organised crime syndicate and continuing unlawful activity rather than the single offence. Its most contested provision makes a confession recorded by a police officer not below the rank of Superintendent of Police admissible in evidence — a direct departure from the ordinary rule that confessions to police are inadmissible. Extended period for filing the charge‑sheet; stringent bail conditions.
Karnataka Control of Organised Crimes Act, 2000 (KCOCA)	Karnataka.	Closely modelled on MCOCA, with the same syndicate‑based architecture and comparable procedural departures.
Uttar Pradesh Control of Organised Crime Act, 2017 (UPCOCA)	Uttar Pradesh.	Follows the MCOCA template. Criticised for extending judicial custody up to one year and police remand up to sixty days, well beyond ordinary criminal procedure.
Jammu and Kashmir Public Safety Act, 1978 (PSA)	Jammu and Kashmir.	A preventive detention statute permitting administrative detention without trial — historically up to two years on grounds relating to the security of the State and up to one year for public order, subsequently amended. Long the subject of domestic and international human rights criticism for detention without charge, and for repeated re‑detention on fresh orders.

The pattern to notice, and it is the examinable point: every one of these statutes achieves its effect by the same three moves — redefining the offence around membership and association rather than a specific act, lengthening pre‑trial detention, and relaxing the rules of evidence or bail in the prosecution’s favour. Once you can name the three moves, you can analyse any special security statute, including ones you have never read.

The 2023 Criminal Law Overhaul

With effect from 1 July 2024, three statutes replaced the colonial‑era codes:

• Bharatiya Nyaya Sanhita, 2023 (BNS) replaced the Indian Penal Code, 1860.
• Bharatiya Nagarik Suraksha Sanhita, 2023 (BNSS) replaced the Code of Criminal Procedure, 1973.
• Bharatiya Sakshya Adhiniyam, 2023 (BSA) replaced the Indian Evidence Act, 1872.

Terrorism enters the general criminal law — Section 113 BNS

The structurally significant change for this paper is that Section 113 of the BNS defines and punishes the “terrorist act” within the general criminal law. A terrorist act is one done with intent to threaten, or likely to threaten, the unity, integrity, sovereignty, security or economic security of India, or to strike terror in the people, by means of explosives, firearms, poisonous or hazardous substances, or by other means, causing death, injury, damage to property or disruption of essential services or supplies. Punishment ranges up to death or imprisonment for life where the act results in death, and otherwise from a minimum term to life imprisonment.

THE DUAL‑TRACK PROBLEM — THE SHARPEST CRITIQUE AVAILABLE India now has two parallel definitions of terrorism: Section 15 of the UAPA and Section 113 of the BNS. They cover substantially the same conduct, but they carry radically different procedural consequences — the UAPA brings a 180‑day charge‑sheet period and the Section 43D(5) bail bar, while a BNS prosecution follows ordinary BNSS timelines and ordinary bail principles. Which track applies is not determined by the nature of the act. It is determined by an officer’s decision — under the scheme adopted, an officer not below a specified senior rank decides whether a given case is registered under the BNS or the UAPA. Why this matters constitutionally. Two persons who have done the same thing may face materially different periods of pre‑trial detention and materially different prospects of bail, on the basis of an unreviewed administrative choice. That is a serious Article 14 problem — equal treatment — and it is the strongest single line of criticism of the 2023 overhaul from an internal security perspective. The counter‑argument. The Government’s position is that placing terrorism in the general criminal law normalises the offence, allows ordinary courts and ordinary police to handle cases that do not require the full special‑law apparatus, and reserves the UAPA for organised and externally linked conspiracies. On this view the dual track is calibration, not arbitrariness — but calibration without published criteria is indistinguishable from discretion.

Procedural changes bearing on security cases

• Police custody may be sought in instalments across the initial period of detention rather than only in the first fifteen days — a significant enlargement of investigative power, and one that applies to all serious offences.
• Trial in absentia for proclaimed offenders is now expressly provided, which has direct application to terrorism accused who have absconded abroad and to designated individuals beyond Indian jurisdiction.
• Mandatory forensic examination for offences carrying seven years or more, together with videography of search and seizure — capacity‑dependent, but in principle a shift from confession‑led to evidence‑led investigation.
• Zero FIR, electronic FIR and prescribed timelines for the completion of investigation and delivery of judgment.
• Under the BSA, electronic and digital records are admissible as primary evidence — of obvious importance to cyber and terror‑finance prosecutions.

From Sedition to Section 152 BNS

Section 124A of the IPC punished any person who, by words, signs or visible representation, brought or attempted to bring into hatred or contempt, or excited or attempted to excite disaffection towards the Government established by law in India.

In Kedar Nath Singh v. State of Bihar (1962), the Supreme Court upheld the provision but read it down, holding that only speech amounting to incitement to violence or a tendency to create public disorder falls within it — mere criticism of government, however strong, does not.

In S. G. Vombatkere v. Union of India (May 2022), the Supreme Court took the extraordinary step of keeping Section 124A in abeyance, directing that no fresh FIRs be registered, that pending investigations be held, and that no coercive steps be taken under it, pending re‑examination of the provision.

The BNS omits Section 124A and the word “sedition” altogether. In its place stands Section 152 — acts endangering the sovereignty, unity and integrity of India — which punishes a person who purposely or knowingly, by words spoken or written, by signs, by visible representation, by electronic communication, by use of financial means, or otherwise, excites or attempts to excite secession, armed rebellion or subversive activities, or encourages feelings of separatist activities, or endangers the sovereignty, unity and integrity of India. The punishment is imprisonment for seven years to life, with fine.

What actually changed

Element	S.124A IPC	S.152 BNS
Protected object	The Government established by law — an institution.	The sovereignty, unity and integrity of India — the State itself.
Conduct	Bringing into hatred or contempt; exciting disaffection.	Exciting secession, armed rebellion, subversive activities; encouraging feelings of separatist activities; endangering sovereignty, unity and integrity.
Means	Words, signs, visible representation.	Words, signs, visible representation, electronic communication, use of financial means, or otherwise.
Mental element	Read down by Kedar Nath to require incitement to violence or tendency to public disorder.	Requires acting “purposely or knowingly” — but the Kedar Nath gloss does not automatically carry over to a differently worded provision.
Punishment	Life imprisonment, or up to three years, with fine.	Seven years to life, with fine — a substantially higher minimum.

The honest assessment. The change is genuinely double‑edged.

• On one side, shifting the protected object from the Government to the State is a real conceptual improvement: criticism of a ruling administration is no longer, on the face of the text, the target.
• On the other side, “subversive activities” and “feelings of separatist activities” are undefined and are at least as elastic as “disaffection” ever was; the express inclusion of electronic communication and financial means widens the reach considerably; the minimum sentence has risen sharply; and the judicial narrowing achieved in Kedar Nath over sixty years attaches to language that no longer exists.

Many commentators therefore argue that Section 152 is potentially wider in operation than the provision it replaced, notwithstanding the disappearance of the word “sedition”.

CURRENT POSITION — READ THIS CAREFULLY BEFORE WRITING ON SEDITION The constitutional challenge is unresolved. In September 2023 a three‑Judge Bench referred the challenge to Section 124A to a larger Bench; that Bench has not yet been constituted. Challenges to Section 152 BNS have been tagged with the Vombatkere petition. The abeyance has been qualified. In May 2026 the Supreme Court clarified that proceedings involving sedition charges under Section 124A may continue where the accused has no objection — typically where an accused or convict wishes his appeal or trial decided in order to obtain acquittal or closure rather than remain in indefinite limbo. The blanket freeze on new FIRs and coercive action was not thereby abandoned. What this means for an answer. Do not write that “sedition has been abolished”. The accurate statement is: Section 124A stands omitted from the new criminal code; a differently framed and arguably wider offence has replaced it in Section 152 BNS; the constitutional validity of both is pending before the Supreme Court; and pending proceedings under the old section continue in qualified circumstances.

Human Rights, the NHRC and the Security–Liberty Balance

The Protection of Human Rights Act, 1993 established the National Human Rights Commission and the State Commissions. In internal security matters, however, the Commission’s powers are at their weakest precisely where the risk is greatest.

SECTION 19 — THE PROVISION TO REMEMBER Where a complaint alleges violation of human rights by members of the armed forces, the Commission may only seek a report from the Central Government. On receipt of that report it may either not proceed with the complaint, or make recommendations to that Government. It cannot conduct an independent investigation, summon witnesses or compel evidence, as it can in an ordinary case. The definitional sting. “Armed forces” is defined to include the naval, military and air forces and any other armed forces of the Union — which has been read to bring the Central Armed Police Forces within the restriction. The CAPFs are the forces most extensively deployed in internal security duties. The Commission’s weakest jurisdiction therefore applies to the forces most likely to generate complaints. The Central Government must inform the Commission of action taken on its recommendations within three months, and the Commission publishes its report with the recommendations and the action taken. But the recommendations are not binding, and there is no mechanism to compel compliance.

Two further constraints matter. The Commission is generally barred from inquiring into a matter after one year from the date of the act complained of — a serious limitation where victims in conflict areas may be unable to approach it promptly. And its recommendations, including on compensation, depend for effect entirely on executive acceptance.

The judicial safeguards that do bite

• D. K. Basu v. State of West Bengal (1997) laid down mandatory requirements for arrest and detention — identification of arresting officers, an arrest memo attested by a witness and countersigned by the arrestee, information to a relative or friend, medical examination, and entry in a police diary — with departmental action and contempt for breach. Substantially incorporated into statute.
• People’s Union for Civil Liberties v. State of Maharashtra (2014) laid down binding guidelines for investigating deaths in police encounters, including registration of an FIR, independent investigation by a different agency or unit, magisterial inquiry, and prompt reporting to the NHRC.
• EEVFAM (2016), examined above, applies the same logic to deaths caused by the armed forces in disturbed areas.

The pattern is clear and worth stating: the accountability that exists in Indian internal security law has been built principally by the courts, not by the legislature. Parliament has enacted the special powers; the Supreme Court has supplied most of the constraints. That is a fragile arrangement, because judicial guidelines depend on individual litigants with the resources to reach the Court, and enforcement of guidelines is notoriously weaker than enforcement of statute.

Critical Assessment

What the legal architecture gets right. India has not governed internal security by executive fiat. Every power is traceable to a statute; every statute is subject to judicial review; and the courts have used that jurisdiction actively — upholding AFSPA but binding it with conditions in Naga People’s Movement, denying absolute immunity in EEVFAM, restoring the constitutional bail jurisdiction in K. A. Najeeb, freezing an entire penal provision in Vombatkere. Compared with the practice of many states facing comparable insurgency, this is a substantial achievement and should be acknowledged before it is criticised.

Where it fails. Five defects recur:

• The exception has become the norm. TADA and POTA had sunset clauses and were allowed to lapse. Their content migrated into the UAPA, which has none. Emergency powers acquired permanence by changing address.
• Process has become punishment. A 1.5–4 per cent conviction rate combined with 80‑plus per cent pendency and a restrictive bail standard means the sanction is imposed by delay rather than by adjudication.
• Accountability is executive‑gated. Section 6 of AFSPA makes prosecution depend on central sanction; Section 19 of the Protection of Human Rights Act makes investigation of the armed forces depend on a central report; UAPA de‑notification is decided by a committee the Central Government constitutes. In each case the body that exercised the power controls the review of it.
• Definitions are elastic. “Unlawful activity”, “subversive activities”, “feelings of separatist activities”, “disturbed or dangerous condition” — none is defined with the precision that criminal law ordinarily requires, and vague criminal law transfers discretion from the legislature to the police officer.
• Duplication creates arbitrariness. Two definitions of terrorism, two procedural regimes, and no published criterion for choosing between them.

CURRENT STATUS (AS OF JULY 2026) UAPA designations: following the amendment of 2019, a total of 80 individuals have been designated as terrorists under Section 35 and listed in the Fourth Schedule. The most recent tranche of 23 individuals — 17 Pakistani nationals and 6 Indian nationals, linked to Lashkar‑e‑Taiba, Jaish‑e‑Mohammed, Jamaat‑ud‑Dawa, Harkat‑ul‑Mujahideen and The Resistance Front — was notified on 4 July 2026. Constitutional challenges to the 2019 amendment remain pending. Sedition: the challenge to Section 124A stands referred to a larger Bench, which has not yet been constituted; challenges to Section 152 BNS are tagged with it. In May 2026 the Court clarified that pending Section 124A proceedings may continue where the accused does not object. AFSPA: in force fully in about 31 districts and partially in about 12 districts across Assam, Nagaland, Manipur and Arunachal Pradesh, extended in six‑month cycles. Withdrawn entirely from Tripura (2015), Meghalaya (2018), Mizoram, and from 23 districts of Assam. The new codes: the BNS, BNSS and BSA have been in force since 1 July 2024, so a body of interpretation is only now developing. Several provisions, including Section 152, are under constitutional challenge. Committee recommendations: the Jeevan Reddy recommendation to repeal AFSPA and relocate its powers into the UAPA with safeguards, made in 2005, remains unimplemented.

SECTION TAKEAWAY India’s security statutes work by three consistent departures from ordinary criminal law: they define offences around association rather than act, they extend detention before charge and restrict bail, and they route accountability through the executive rather than the court. Each departure is defensible in isolation and each is dangerous in accumulation. The result is a body of law in which the exception has outlived every mechanism designed to end it — TADA and POTA died on their sunset clauses, but their substance lives on in a statute that has none. The reform agenda is not to disarm the State. It is to restore the two features the framers assumed would discipline exceptional power: a time limit, and a reviewer who is not the same person who exercised the power.