Jan Gallemí Albás es estudiante del Máster en Seguridad Internacional en Sciences Po, en París, donde se especializa en inteligencia y riesgos globales.
Actualmente combina su formación académica con su labor como Young European Ambassador y su colaboración con centros de análisis como Global Affairs and Strategic Studies y el Centro de Pensamiento Naval, centrando su trabajo en cuestiones de seguridad internacional. Se graduó en el doble grado de Relaciones Internacionales e Historia por la Universidad de Navarra, periodo durante el cual amplió su perspectiva internacional con estancias en Strathmore University y la Universidad Hebrea de Jerusalén. Ha desarrollado experiencia institucional a través de prácticas en la Armada, tanto en la Escuela de Guerra Naval como en la Dirección de Asuntos Económicos, así como en el medio África Mundi. Además, ha sido representante juvenil de Kinderenergy ante Naciones Unidas en Nueva York.
Es autor de más de una veintena de artículos en el ámbito de las relaciones internacionales.

The integration of artificial intelligence (AI) into military intelligence has become a defining feature of 21st-century security environments, particularly within a highly competitive international system characterized by rapid technological change and exponential data growth. Intelligence operations increasingly depend on the ability to exploit vast volumes of multi-source information, where the capacity to collect, process, and analyze data has become a decisive factor for operational effectiveness. In this context, AI has emerged not merely as a supporting tool but as a critical enabler of modern intelligence, especially within hybrid environments where information operations and cognitive competition are central.

This memorandum analyzes two key documents. The first examines the Artificial Intelligence for the Intelligence Community (AIM) Initiative developed by the United States Intelligence Community. The second is an academic study published by the Center for Intelligence and Security Studies at the University of the Bundeswehr Munich titled “The Use of Artificial Intelligence in Military Intelligence: An Experimental Investigation of Added Value in the Analysis Process.” This study provides a crucial scientific foundation by empirically demonstrating that the use of AI tools, specifically semantic search, automatic summarization, and Named Entity Recognition (NER), enables analysts to obtain significantly more accurate assessments and probability estimates closer to expert judgment, especially under time-pressured conditions. The objective is to correlate the ideas presented in these works with the strategic and operational vision of AI reflected in Spanish defence literature, particularly publications by the Spanish Ministry of Defence. This comparison is especially relevant given the relatively limited body of Spanish-language academic research on AI applications in military intelligence.

The AIM Initiative was developed in response to a structural challenge: the growing gap between the volume of data collected by intelligence systems and the capacity of human analysts to process it effectively. As the initiative itself states, “closing the gap between decisions and data collection is a top priority for the Intelligence Community.” This ambition is reflected in the well-known statement attributed to former Principal Deputy Director of National Intelligence Sue Gordon: “If it is knowable, and it is important, then we know it.” At its core, the initiative seeks to secure a sustained strategic information advantage through the accelerated adoption of Artificial Intelligence, Automation, and Augmentation (AAA).

Within this framework, AI is conceived not as a replacement for human analysts but as a force multiplier for cognitive and analytical performance. Machine learning systems are expected to assist in pattern recognition, anomaly detection, and the synthesis of heterogeneous data sets that exceed human processing limits. This approach directly addresses the phenomenon of “data overload,” in which intelligence collection capabilities have outpaced analytical capacity

The Bundeswehr Munich academic study adds depth to the analysis by identifying the limits of AI’s “added value.” While it confirms a demonstrated increase in the speed and accuracy of analysis, it notes that AI provides clear benefits in factual and concise tasks, but its usefulness declines sharply in tasks requiring complex argumentation or the analysis of ambiguous and contradictory information. Likewise, the experiment revealed a “confidence paradox”: although analysts’ performance improved with AI, their subjective trust in the sources did not increase, underscoring the urgency of developing Explainable AI (XAI) to prevent systems from being perceived as inscrutable “black boxes.”

The AIM Initiative defines four principal investment priorities aimed at transforming the analytical enterprise. First, the development of a robust digital infrastructure capable of supporting large-scale data integration and computation. Second, the rapid adoption of commercial and open-source AI solutions to enable short-term operational gains. Third, the advancement of multimodal AI systems able to integrate outputs from disciplines such as GEOINT, SIGINT, and OSINT. Finally, long-term research into machine reasoning and sense-making capabilities designed to support complex analytical judgment. Taken together, these priorities reveal a dual approach that combines immediate operational enhancement with long-term technological transformation.

Comparable trends can be identified in Spanish defence thinking. Analysts such as Ruiz highlight similar objectives in initiatives like the Sistema de Mando y Control Nacional (SC2N), which seeks to integrate multiple intelligence subsystems and enable real-time decision support. Spanish doctrinal publications likewise emphasize the growing relevance of data science techniques for identifying relational patterns across large datasets, particularly in counterterrorism and the analysis of complex networks.

Spain has formalized this vision through the «Strategy for the development, implementation and use of AI in the Ministry of Defense» (Resolution 11197/2023). The Spanish case is distinguished by its focus on Western humanism, arguing that technology should not replace military judgment or moral responsibility. Key components of the Spanish strategy include three important elements. To start with, the Centro de Referencia de IA (CRIA), which is included in the  Mando Conjunto del Ciberespacio (MCCE), which leads design of solutions for the strategic and operational levels, focusing on the planning and conduct of operations. At the same time, another element is the Combat Cloud, Envisioned as a “system of systems” for transparent data sharing in the battlespace, allowing AI to act as a central processor in multi-domain operations. But the final element is ethical. As stated by a professor from Jaen University, Spain explicitly rejects the use of lethal autonomous weapons systems (SAAL), insisting that any system must allow human supervision that guarantees traceability and accountability in each phase of the action.

A key divergence lies in the balance between innovation and regulation. The AIM Initiative, while acknowledging the importance of trust and security, prioritizes speed, adaptability, and the preservation of technological superiority, explicitly warning against bureaucratic constraints that may hinder innovation. This approach is shaped by strategic competition with technologically advanced adversaries such as China and Russia. In contrast, Spanish defence literature, reflecting the broader European regulatory environment, places greater emphasis on legal oversight, ethical safeguards, and human control. Concepts such as transparency, accountability, and algorithmic responsibility are central, aligning closely with European Union frameworks for trustworthy AI. Spanish doctrine emphasizes that technology does not change essential military values, but rather modifies the competencies necessary to exercise them under the principle that moral reason must always precede technical reasons.

Both perspectives also converge on a set of shared technical challenges. Among the most significant is model degradation caused by concept drift, whereby machine learning systems lose accuracy as operational environments evolve. In intelligence contexts, this risk is compounded by adversarial interference, including disinformation and data manipulation intended to mislead analytical systems. Ensuring data integrity, model robustness, and rigorous validation processes therefore becomes essential for the reliable deployment of AI in intelligence operations.

International cooperation constitutes another critical dimension. The United States emphasizes deep intelligence integration within alliances such as the Five Eyes, facilitating interoperability and shared technological development. Spain, by contrast, prioritizes cooperation within NATO and the European Union hosting forums such as the REAIM Summit on Responsible AI and collaborating with the NATO DARB working group for the ethical use of data. Spain also promotes civil-military collaboration through initiatives like the Consejo Asesor de Innovación de la Defensa (CAID). These frameworks reflect differing strategic cultures but a shared recognition that AI development in intelligence is inherently collaborative.

In conclusion, the AIM Initiative represents a pragmatic and strategically driven effort to maintain analytical superiority in an era defined by data saturation. Its emphasis on speed, scalability, and technological dominance contrasts with the more regulation-oriented approach found in Spanish defence literature, which prioritizes ethical governance and institutional accountability. However, the academic evidence from the Bundeswehr study and the rigor of the Spanish ethical framework demonstrate that the effectiveness of military intelligence does not reside only in computing power, but in the quality of man-machine interaction. Rather than representing mutually exclusive models, these approaches can be understood as complementary: one optimized for competitive advantage in high-intensity strategic environments, the other for legitimacy and sustainability within democratic systems. The central challenge moving forward will lie in reconciling these priorities, ensuring that the pursuit of technological superiority does not undermine the very legal and ethical frameworks that intelligence institutions are ultimately meant to defend.

Declaration on the use of AI:

The preparation of this work involved an initial close reading of the selected sources, including key doctrinal materials on the Spanish defence framework. Based on this review, an initial draft was developed by the author.

Subsequently, AI-assisted tools were used in a limited and supervised manner, in accordance with institutional guidelines on AI use. NotebookLM was employed for Type 2 assistance (support with the organisation and cross-referencing of source material), facilitating the identification of relevant thematic connections. ChatGPT was used exclusively for Type 3 assistance (language editing and stylistic refinement), with the aim of improving clarity and stylistic consistency.

All analytical interpretations, structuring decisions, and substantive arguments remain the sole responsibility of the author.

Bibliography:

Arteaga Martín, F. (2019). «Contexto estratégico de la inteligencia artificial». En: La Inteligencia Artificial aplicada a la Defensa. (Documentos de Seguridad y Defensa, 79), pp. 99-126. Instituto Español de Estudios Estratégicos (IEEE) – CESEDEN.

Ballenilla y García de Gamarra, M. (2026). El mando ante el algoritmo: Inteligencia artificial, doctrina, inteligencia y disuasión. Documento de Análisis IEEE 35/2026. Instituto Español de Estudios Estratégicos (IEEE).

CASEDE. (2021). Carpeta Inteligencia Artificial: Documentos y Publicaciones. Colectivo de Análisis de la Seguridad con Democracia A. C..

España. Ministerio de Defensa. (2025). Resolución de la Asamblea General de Naciones Unidas (AGNU) 79/239 de 31 de diciembre de 2024. La Inteligencia Artificial en el ámbito militar y sus implicaciones para la paz y la seguridad internacional. Abril 2025.

Garat González, J. M. (2024). La inteligencia artificial como factor de transformación de las operaciones militares en el nivel operacional. Documento de Opinión IEEE 12/2024. Ministerio de Defensa / IEEE.

Indra. (2021, 18 de febrero). Indra, reconocida por su aportación al desarrollo científico de la Inteligencia Artificial en la industria de defensa y seguridad. Comunicado de prensa.

Millán Martínez, J. M. (2025). Conferencia inaugural: Inteligencia Artificial y defensa. Revista de Estudios Jurídicos, nº 25. Universidad de Jaén.

Moliner González, J. A. (2018). «Desafíos éticos en el uso militar de la inteligencia artificial». En: La Inteligencia Artificial aplicada a la Defensa. (Documentos de Seguridad y Defensa, 79). IEEE-CESEDEN.

Moliner González, J. A. (s.f.). Valores éticos frente al uso de sistemas de armas letales autónomas. Academia de las Ciencias y las Artes Militares / Ministerio de Defensa.

Nitzl, C., Cyran, A., Krstanovic, S., & Borghoff, U. M. (2025). The use of artificial intelligence in military intelligence: an experimental investigation of added value in the analysis process. Frontiers in Human Dynamics, 7:1540450. doi: 10.3389/fhumd.2025.1540450.

Office of the Director of National Intelligence (ODNI). (2019). The AIM Initiative: Augmenting Intelligence using Machines. United States Intelligence Community.

Roca Rivero, F. J. (2023). La era de la Inteligencia Artificial ya está aquí. Ministerio de Defensa de España. Vicealmirante Comandante del Mando Conjunto del Ciberespacio.

Roldán Tudela, J. M. (Presidente). (2018). La inteligencia artificial aplicada a la defensa. Documentos de Seguridad y Defensa 79. Instituto Español de Estudios Estratégicos (IEEE) – CESEDEN.

Talavera Rodríguez, E. (2025). Impacto de la regulación en la aplicación de Inteligencia Artificial en defensa y seguridad. Documento de Opinión IEEE 70/2025. Instituto Español de Estudios Estratégicos (IEEE).