Prime Minister Modi’s Security Doctrine Builds On Lessons From Operation Blue Star

41 years separate Operation Blue Star from Operation Sindoor.

In June 1984, the Indian Army went into the Golden Temple in Amritsar with no reliable assessment of what was waiting inside, and with a training plan that had already leaked to the other side. In May 2025, India struck nine terrorist facilities across Pakistan and PoK in under thirty minutes — targets that had been surveilled, mapped, and sitting in intelligence files for years before the political trigger arrived.

That distance is not accidental. It is four decades of institutional rebuilding, forced by the failures of 1984, and the foundation on which Modi’s security doctrine has been built — not invented by his government, but carried furthest by it.

The Movement The State Failed

To Read

Bhindranwale’s movement did not ambush the Indian state. It grew in front of it. His rise through the late 1970s was aided, at various points, by Congress politicians who found him a convenient instrument against the Akali Dal. By July 1982, he had been invited into the Golden Temple complex itself by Akali Dal president Harchand Singh Longowal, and the government arrested him that September, only to release him two days later, unable to make a case because nobody had built the networks to make one.

What happened inside the complex over the following two years was both visible and, as far as the state’s intelligence apparatus was concerned, essentially unread. Major General Shahbeg Singh, a former Indian Army officer who had been court-martialled and had since redirected his knowledge of military tactics towards the other side, was fortifying the Akal Takht with professional deliberateness. Seventeen houses surrounding the complex were occupied as forward positions, some 800 yards out, all in wireless contact with a command centre inside.

RPG launchers with armour-piercing ammunition, automatic weapons, prepared lines of fire through the lanes — it was a defensive structure built methodically over months. The agencies responsible for monitoring the situation did not produce a picture that reflected what was being constructed.

The Army, to its credit, had been planning long enough to build a full-scale replica of the temple complex at Chakrata Cantonment in the Doon Valley for commando training. That plan was scrapped when it leaked to the militants. The operational plan for what would become Blue Star had reached the adversary before a single soldier moved.

Barefoot At The Threshold

On the night of June 5, 1984, as soldiers from the 1st Para Regiment prepared to enter Harmandir Sahib, many of them took off their boots. Soldiers going to war inside a shrine, removing their shoes at the door.

It is a detail that gets lost in the military and political accounting of Blue Star, but it says a lot about the impossibility of the operation. The Army had been sent to conduct a combat operation inside the holiest site of a major religion, on the martyrdom anniversary of Guru Arjan Dev, with thousands of civilian pilgrims trapped inside who had nothing to do with the militancy. The force tasked with the job was a conventional infantry formation, trained for open-terrain warfare, handed an intelligence picture that significantly underestimated what Shahbeg Singh had built.

The 1st Para Regiment entered through the main gateway and walked into prepared kill zones. Tanks that had been brought for elevated fire support had to be turned directly against the Akal Takht to break the resistance. The Army and its agencies had grossly under-assessed the level of resistance inside the temple. The government’s own White Paper recorded 554 combined militant and civilian deaths.

The Akal Takht, the temporal seat of Sikh religious authority, was left in ruins. Prime Minister Indira Gandhi was assassinated on October 31, 1984, by two of her Sikh bodyguards. The Intelligence Bureau had reportedly flagged concerns about Sikh personnel in her security detail in the aftermath of Blue Star. No changes were made to her protective arrangement. She was killed in her own garden. The operation’s tactical objective of eliminating Bhindranwale was achieved. Everything around it collapsed.

The Rebuilding

The National Security Guard was constituted on October 16, 1984, directly from the Blue Star analysis. The question it was meant to answer was: who do you call when the situation requires close-quarters counter-terrorism in sensitive environments and you cannot send an armoured infantry battalion? The NSG Act was passed in 1986, with the force modelled on the SAS and GSG-9.

The Rashtriya Rifles followed in 1990 for Jammu and Kashmir, built around long deployments because infantry units rotating through a conflict zone on short tenures cannot build the human intelligence networks that counter-insurgency depends on. By the early 1990s, this model, combined with intelligence penetration of militant networks in Punjab, had begun producing results that force alone had not.

By 2014, India had a meaningfully different security architecture from 1984. Ajit Doval, Modi’s choice as National Security Adviser, had spent years as an IB field officer in Punjab during precisely that rebuilding period. He had seen what intelligence failure looked like and what patient intelligence work produced. His appointment put someone at the centre of security decision-making who carried that experience as professional memory, not textbook history.

What Changed Under Modi

Previous governments had built the capability. Modi’s government decided to use it differently — openly, offensively, and with the stated intention of changing the terms of engagement with Pakistan-backed terrorism.

The shift was made explicit in how the 2016 surgical strikes were handled. On the night of September 28–29, following the attack on the Uri Army base that killed 19 soldiers, Para Special Forces crossed the Line of Control and struck four militant launch pads. This was not the first time India had conducted cross-LoC operations, but it was the first time a government announced it publicly the following morning, by name, with an official briefing by the DGMO.

The Northern Command had been running human sources inside Hizbul Mujahideen, developing ground-level intelligence on Pakistani camp layouts and movement corridors. The targets had been known for years. What had changed was the political decision to act on them and then say so.

Balakot, on February 26, 2019, was a larger step. Twenty Mirage 2000 jets struck a Jaish-e-Mohammed training facility in Pakistan’s Khyber Pakhtunkhwa province. It was the first Indian air strike inside Pakistani territory since 1971. India’s official description was “intelligence-led, non-military and pre-emptive.” The camp had been under surveillance for considerably longer than the weeks following Pulwama.

Operation Sindoor, on May 6–7, 2025, was the most extensive operation. Nine sites were struck in Pakistan and Pakistan-administered Kashmir in under thirty minutes. Jaish-e-Mohammed’s infrastructure in Bahawalpur, and Lashkar-e-Taiba and Hizbul Mujahideen facilities in Muzaffarabad and Kotli, were the targets.

No Pakistani military assets were targeted — a deliberate boundary. The operation followed the Pahalgam attack of April 22, in which 26 tourists were killed in the Baisaran Valley by Lashkar-e-Taiba gunmen. But those nine targets had not been developed in the two weeks between Pahalgam and the strikes. India knew where to hit because the intelligence work had been accumulating for years. The speed of the response was a function of preparation, not urgency.

In the hours after Sindoor, India briefed the media domestically and internationally before Pakistan’s account of events could settle into the public record. The National Security Council Secretariat coordinated the narrative alongside the operation itself. Compare that to 1984, when the Army struck and the government spent the following months losing the political argument entirely, at home and abroad.

Where The Gaps Remain

This account would be incomplete without Pahalgam and Burhan Wani, because both sit inside the same government’s tenure and neither reflects well on intelligence performance.

Wani had been building his following openly on Facebook and YouTube since 2015 — photographs in camouflage, videos from the forest, and direct recruitment appeals to Kashmiri youth. He was a known militant with a bounty on his head. When he was killed in Anantnag on July 8, 2016, the Valley went into 53 consecutive days of curfew. More than 96 people lost their lives and over 15,000 were injured. The open-source picture of Wani’s reach and influence was there. Reading it as an operational intelligence problem, and acting on that reading before the trigger was pulled, was not.

Pahalgam, in April 2025, was worse in terms of civilian lives. Intelligence had reportedly flagged the movement of hybrid militants into south Kashmir in the weeks before the attack. Pahalgam and Sindoor happened within weeks of each other, under the same government. The doctrine is real. So are the gaps in applying it consistently.

Blue Star’s Lasting Relevance

The failures of June 1984 were the product of an institution that lacked the right capabilities, had not gathered the right intelligence, and had not thought through what it meant to conduct a security operation inside a religious space that was politically combustible, where conventional military logic was the wrong instrument entirely.

India has spent forty years correcting that. The Modi government has been the most visible in deploying the corrected version, and the most willing to say openly when it does. The doctrine that has emerged is built on the premise that intelligence drives action, not the other way around.

Blue Star is where India learned what it cost to operate without that in place. Pahalgam is a reminder that the lesson requires active maintenance. The assumption of stability, when it replaces sustained vigilance, has always produced the same result — and 1984 is the clearest proof of what that result looks like at its worst.