From “Blinking Red” to AMIA: An Emerging Standard of Intelligence Failure Under the Right to Life

On 1 September 2004, more than a thousand people were taken hostage in School No. 1 in Beslán, North Ossetia. Three days later, at least 334 of them — including 186 children — were dead. On 18 July 1994, a vehicle laden with explosives detonated in front of the headquarters of the AMIA — the Asociación Mutual Israelita Argentina, the central institution of Argentina’s Jewish community — in Buenos Aires, killing 85 people and injuring more than 150. In April 2017, the European Court of Human Rights held Russia liable under Article 2 ECHR in Tagayeva and Others v. Russia. In January 2024, the Inter-American Court did the same to Argentina under Articles 4(1) and 5 ACHR in Asociación Civil Memoria Activa v. Argentina, Series C No. 516 (26 January 2024) (the AMIA case). The two judgments reached the same diagnosis along independent doctrinal paths; and, in a move that has so far attracted little attention, the Inter-American Court expressly invoked the Strasbourg Court’s §482 of Tagayeva in §125 of AMIA, as it entered for the first time the specific terrain of preventive intelligence failure in counter-terrorism.

This post argues that what these two cases articulate, read against the extra-jurisdictional convergence offered by the 9/11 Commission Report, is the slow emergence of an autonomous doctrinal category within the positive duty to protect life: an intelligence-failure standard. This standard is identifiable by a qualified activation threshold, a structural evidentiary asymmetry, and an organizational locus for breach. The category is not yet self-described by either court, but the cross-citation in AMIA §125 signals that the Inter-American system has chosen to locate itself, on this point, in Strasbourg’s slipstream. It is the supranational counterpart, built from within the right to life, of the negligence regime that intelligence-law scholarship has pursued from outside it.

The European Anchor: Tagayeva’s Three-Layered Operational Duty

The operational duty under Article 2 has its modern root in Osman v. United Kingdom (1998, §116): real and immediate risk to an identified individual, plus reasonably available measures the State failed to take. Tagayeva extends and refines that test along three layers.

First, knowledge. The Court works by accumulation (Tagayeva §§484–489): internal MVD/FSB directives of July–August 2004; the historical pattern of analogous attacks (Budennovsk 1995, Kizliar 1996, Dubrovka 2002); and expert evidence that “elementary information should have been available to the competent authorities from covert sources and intelligence operations” (§486). The duty includes the active generation of reasonably obtainable intelligence, not merely reaction to what is received.

Second, evidentiary asymmetry. In §492 the Court formalises what was previously implicit: “the police, who have access to information and intelligence not available to the general public, will usually be in the best position” to take operational decisions. Structurally, this places a justificatory burden on the State where the intelligence picture is opaque to victims. The implications of this move for the evidentiary regime in intelligence-related cases have not yet been fully drawn out.

Third, organisational content. The violation crystallises in the absence of “a single sufficiently high-level structure responsible for the handling of the situation, evaluating and allocating resources, creating a defence for the vulnerable target group and ensuring effective containment of the threat and communication with the field teams” (§491). What Tagayeva condemns is not a missing (positive) measure, but a missing architecture.

The Cross-Citation Event: AMIA §125 and the Inter-American Importation

AMIA §125, in its discussion of the prevention duty in counter-terrorism (Section VII.B.2), reproduces Tagayeva §482 as a block citation (note 150):

“Turning to the question of positive obligation, the Court reiterates that Article 2 of the Convention may imply a positive obligation on the authorities to take preventive operational measures to protect an individual whose life is at risk from the criminal acts of another individual. For the Court to find a violation of the positive obligation to protect life, it must be established that the authorities knew, or ought to have known at the time, of the existence of a real and immediate risk to the life of identified individuals from the criminal acts of a third party and that they failed to take measures within the scope of their powers which, judged reasonably, might have been expected to avoid that risk. Such a positive obligation may apply not only to situations concerning the requirement of personal protection of one or more individuals identifiable in advance as the potential target of a lethal act, but also in cases raising the obligation to afford general protection to society.”

The textual choice deserves note. Osman §116 stated the test in disjunctive form (‘an identified individual or individuals’); Tagayeva §482 retained only the plural. AMIA §125 imports that compressed formulation — the version Strasbourg had already worked in the Beslán fact-pattern — rather than the 1998 source. The Inter-American Court thus joins the Strasbourg lineage at its current stage     , not at its origin.

The move is not formally ratio decidendi: the Court’s operative test sits in §118, applied to the AMIA facts in §§128–137, and rests on its own Pueblo Bello lineage, with the five-pronged test of UN Special Rapporteur Ben Emmerson (Report A/HRC/20/14, 2012) explicitly received in §123. But even in obiter, the institutional choice to invoke Strasbourg precisely here, as the Court enters preventive intelligence failure in terrorism for the first time, places the Inter-American system in Strasbourg’s wake on Article 2.

The application of the standard to AMIA’s facts converges with Tagayeva’s grammar. The Court documents three knowledge inputs (§§130–133): Argentina’s own recognition of responsibility in Decreto 812/2005; the 1992 Embassy bombing and the resulting “Sala Independencia” within the SIDE intelligence service; and the two immediate alerts — the W.R.D.S. tip from the Argentine consulate in Milan, and the low-altitude helicopter overflights the day before and the day of the attack, both evaluated with expert input from Prof. Martin Scheinin’s amicus-style report (published as Bonavero Report 1/2022).

The organisational diagnosis mirrors Tagayeva §491 in AMIA §§135–136: the patrol assigned to Pasteur Street had been stationary for three days with its radio out of service; there was no coordination between external and internal security; and the State “did not investigate the information given by W.R.D.S. nor the helicopter overflight”. The Court reads the prevention duty as an obligation of means and not of result (§136) — the Inter-American twin of the Strasbourg formulation reproduced above.

Extra-Jurisdictional Verification: The 9/11 Commission Report

Why does the convergence matter beyond the two systems? Because the United States did not ratify the American Convention, has not accepted the Inter-American Court’s contentious jurisdiction, and is outside the European Convention. Any convergence between the Strasbourg–San José line and the U.S. institutional record is therefore independent: it does not presuppose the doctrinal pattern, it strengthens it.

Chapter 8 of the 9/11 Commission Report — “The System Was Blinking Red” — reconstructs the unprecedented volume of pre-9/11 warning traffic, anchored in the line attributed to DCI George Tenet (“the system was blinking red”). The empirical record (June–July 2001 directives, CSG warnings, FBI advisories, the 6 August PDB “Bin Ladin Determined to Strike in US”) is functionally equivalent to what Tagayeva §§484–486 and AMIA §§130–133 document.

The decisive convergence is organisational. Section 11.4 of the Report (“Management”) delivers, in administrative prose, the precise diagnosis that Tagayeva §491 articulates in convention-rights terms: “The agencies are like a set of specialists in a hospital, each ordering tests, looking for symptoms, and prescribing medications. What is missing is the attending physician who makes sure they work as a team.” The “attending physician” and the “single sufficiently high-level structure” name the same institutional pathology. The remedial coda confirms the diagnosis: the Commission proposed, and the Intelligence Reform and Terrorism Prevention Act of 2004 created, the National Counterterrorism Center and the Director of National Intelligence, the coordinating authority whose absence was identified as causal.

An Emerging Standard: Four Elements for Future Cases

The standard is not yet judicially self-described. The four elements below reconstruct the converging pattern descriptively from the case law, with the qualifications noted in passing where the doctrine still has to catch up.

Element 1 — Available or required intelligence. The State must have had access, before the harmful event, to intelligence capable of generating a common operational picture of the structural threat; or, applying the diligence reasonably expected of a comparable intelligence agency, it must have been in a position to procure it. The “required” prong captures Tagayeva‘s expectation of active generation (§486), not only reactive use.

Element 2 — Convertibility into a specific situational picture. The general threat picture must be convertible, through reasonable diligence, into a specific picture against an identifiable target — individual, vulnerable group, or institution — in determinate coordinates of time, place, and modality. This is the qualified threshold that distinguishes the category from canonical Osman/Pueblo Bello: it excludes genuine operational unpredictability (Finogenov v. Russia, 2011, where the Court accepted the unforeseeability of the Dubrovka rescue plan) while capturing cases where conversion was possible and was not produced.

Element 3 — Absence of organisational architecture of conversion. The standard requires the absence of a single, sufficiently senior authority to centralise threat management; a mechanism of resource evaluation and allocation; an active capacity to build a defence for the identifiable vulnerable group; and an operational communication loop with field teams. Documented dysfunction of a nominally existing apparatus (Pasteur’s stationary patrol, the broken radio) is sufficient.

Element 4 — Materialised harm. The harm must materialise in the convention-protected interest the conversion would have prevented or mitigated. Causation is probabilistic, in line with obligations of means, and the Tagayeva §492 asymmetry operates in the victim’s favour.

Read against Asaf Lubin’s Reasonable Intelligence Agency standard (47 Yale J. Int’l L. 119, 2022) and his Liberty to Spy framework (61 Yale J. Int’l L. 185, 2020), what Tagayeva and AMIA are doing is building, from within the right to life, the supranational counterpart of the negligence regime that intelligence-law scholarship has pursued from the outside. The cross-citation in AMIA §125 is the moment that move became visible.