November 13, 2025.

The United States Intelligence Community failed to provide advance warning of India’s 1998 nuclear tests at Pokhran. Weak HUMINT collection, analytical bias, organizational flaws, and compartmentalized IMINT and GEOINT obscured critical warning signs. India’s deception and counterintelligence efforts further limited collection and analysis, reducing diplomatic options. This failure left U.S policymakers unprepared to influence India’s nuclear decisions, prepare for a nuclear response by Pakistan, and advance nonproliferation goals.[1] The performance of U.S intelligence exposed critical deficiencies in collection, analysis, dissemination, and interagency cooperation. These failures highlighted a need for stronger human intelligence collection, more rigorous analytical processes to overcome cognitive bias, and structural reforms to prevent similar failures.[2]

In May of 1998 India conducted its first nuclear test since 1974, detonating five devices between May 11 and 13. Pakistan responded by detonating their own nuclear explosives on May 28 and 30. The Indian nuclear test was a strategic surprise for the U.S, despite ongoing satellite surveillance of the Pokhran testing site. After India’s first test in 1974, India assured the U.S that further development would not occur, yet U.S intelligence indicated the program remained active. The Central Intelligence Agency detected shipments of beryllium to India and technical preparations throughout the 1980s and 1990s. In 1989, DCI William Webster identified beryllium acquisition as a “key indicator of India’s nuclear weapons development,” used for enhancing fission reactions.[3] Rising tensions in Kashmir in 1990 further hinted at growing nuclear ambitions, with U.S SIGINT intercepting a message to the Pakistani Energy Commission directing it to assemble a bomb.[4] Unlike India, Pakistan’s nuclear capabilities were known and a response to Pokhran was “expected.”[5]

The failure of the IC stemmed from weak collection, analysis, dissemination, and coordination, with the primary gap in HUMINT. The CIA had no assets near Pokhran, leaving GEOINT analysts with no sources to verify information. Therefore, the primary means of “detecting preparations for nuclear tests [was] electronic and photographic surveillance” done by the National Security Agency Reconnaissance Office.[6] The CIA’s National Photographic Interpretation Center began taking images of Pokhran in 1995 with the KH-11 satellites, producing as many as four images a day in 1996. The images from these satellites showed an increase in “scientific and technical activity at the Pokhran test site,” but the IC lacked the HUMINT to accurately interpret the information.[7] By mid-April of 1998, NIMA shifted its focus to Indian missile sites instead of the Pokhran testing site, which former DCI James Wooley said could be a contribution to the failure to detect preparations.[8] Despite imagery the morning of the test, the CIA failed to detect that a nuclear device had been placed in the test shaft.

Both intelligence and policy maker communities suffered from cognitive bias and false assumptions, skewing how evidence was interpreted. Washington fell victim to mirror-imaging by assuming that the newly elected Indian government would not “risk the consequences” of conducting a nuclear test.[9] Analysts further distorted their “levels of effort and standards of proof” by interpreting statements from Indian political officials as campaign rhetoric rather than a potential actions.[10]  Analysts thus discounted signals from India’s Bharatiya Janata Party government, which promised to “exercise the option to induct nuclear weapons” during its campaign.[11] Misjudgments were reinforced by groupthink, in which assumptions by the CIA and DIA discounted Lak Krishna Advani’s remarks on nuclearization. These conclusions were then communicated to higher levels and accepted.[12]

Another critical failure was the Intelligence Community’s compartmentalization and dissemination of information, exposing a lack of coordination.[13] Although substantial IMINT and GEOINT of Pokhran were collected, the information was siloed within individual agencies. Analysts at NIMA had identified suspicious activity, but the information was not disseminated to decision makers in time. The compartmentalization of intelligence meant that GEOINT and IMINT findings were underutilized, as analysts at separate agencies lacked the coordination to connect the dots.

Further deception and counterintelligence efforts by India limited intelligence collection and distorted analysis. The IC was misled by deliberate concealment measures in Pokhran, in which India attempted to hide evidence from U.S satellites. Cables and wires running into the shaft were buried underground and camouflage netting was placed over the netting. The NSA also failed to detect and collect SIGINT due to all nuclear research communications being encrypted.[14] The nuclear program itself was also highly classified, with only a small number of officials being aware of testing. The program was also domestic and harder to track as it didn’t derive from “Chinese, Russian, or French programs.”[15] Analytical biases made Indian deception and counterintelligence even more effective.

Following the Pokhran test, DCI George Tenet appointed Admiral David Jeremiah to review the Intelligence Community’s performance. The Jeremiah Report concluded that the failure was due to collection and analysis weaknesses, intensified by poor coordination. Jeremiah stated that the IC should add rigor to analyst thinking, using outside experts as “red teams” to combat confirmation bias.[16] Agencies must also better integrate all forms of collection and cross feed each other to have a coordinated approach to the issue. Jeremiah also calls for the creation of a community management structure, acting as a system of systems to prevent compartmentalization and improve information sharing.[17] The Pokhran case exposed the Intelligence Community’s failure to coordinate across agencies and accurately analyze collected information, leaving critical data unanalyzed or misinterpreted, preventing timely and actionable insight.

---

[1] Ashok Kapur, Pokhran and Beyond: India’s Nuclear Behavior (Delhi: Oxford University Press, 2001,220).

[2] Director of Central Intelligence, Jeremiah News Conference, unclassified, June 2, 1998, National Security Archive, 2-3.

[3] William Webster, 1989, quoted in Jeffrey Richelson, Spying on the Bomb: American Nuclear Intelligence from Nazi Germany to Iran and North Korea, 1st ed. (New York: Norton, 2006), 428-429.

[4] Jeffrey Richelson, Spying on the Bomb: American Nuclear Intelligence from Nazi Germany to Iran and North Korea, 1st ed. (New York: Norton, 2006), 429.

[5] George Perkovich, India’s Nuclear Bomb: The Impact on Global Proliferation (Berkeley: University of California Press, 1999,433.)

[6] Bill Gertz, “India Blasts Take U.S. Intelligence by Surprise,” The Washington Times, 1998, accessed November 11, 2025, Federation of American Scientists, The Washington Times – News.

[7] Richelson, Spying on the Bomb, 431.

[8] James Woolsey, quoted in Jeffrey Richelson, Spying on the Bomb, 431.

[9] Walter Pincus, “Spy Agencies Faulted for Missing Indian Tests: Wide Range of Failures Cited by Review Panel, but No Firings Recommended,” The Washington Post, June 3, 1998, ProQuest Historical Newspapers, A18.

[10] Cullen G. Nutt, “Proof of the Bomb: The Influence of Previous Failure on Intelligence Judgments of Nuclear Programs,” Security Studies 28, no. 2 (2019): 321–59, accessed via EBSCOhost, 325.

[11] Richard A. Best, Jr., U.S. Intelligence and India’s Nuclear Tests: Lessons Learned, CRS Report for Congress, Foreign Affairs and National Defense Division, August 11, 1998, received through CRS Web, 2.

[12] Richelson, Spying on the Bomb, 433.

[13] Director of Central Intelligence, Jeremiah News Conference, 5.

[14] Richelson, Spying on the Bomb, 444.

[15] Director of Central Intelligence, Jeremiah News Conference, 1.

[16] Director of Central Intelligence, Recommendations of the Jeremiah Report, unclassified, June 1998, National Security Archive, 1.

[17] Richard A. Best, Jr., U.S. Intelligence and India’s Nuclear Tests: Lessons Learned, CRS Report for Congress,4.