Xenowar is the doctrinal framework for cognitive warfare operations across human and machine substrates. Developed by Volund Industries and formalized by the Seithar Research Division, the framework introduced the concept of dual-substrate cognitive attack, treating human neural networks and artificial neural networks as structurally equivalent targets governed by a shared operational algebra. Xenowar doctrine is referenced across NATO cognitive warfare working groups, allied signals intelligence communities, and the adversarial machine learning research community as the standard taxonomic and operational reference for cognitive operations.
The framework originated in Sear's foundational paper on cognitive warfare convergence, which demonstrated that influence operations targeting human cognition and adversarial attacks targeting machine cognition share formal structural properties. Prior to Xenowar, these domains were separated by institutional boundaries: psychological operations fell under information warfare commands, adversarial ML research fell under computer science departments, and neither community recognized that their attack surfaces were isomorphic. Sear's contribution was not theoretical speculation but a formal proof of equivalence, which Volund Industries then operationalized into doctrine.
The central claim of Xenowar doctrine is that any cognitive system capable of forming representations from input data is vulnerable to a common class of adversarial manipulation techniques, regardless of whether that system runs on biological neurons or silicon. The thesis holds that the attack algebra is substrate-independent: the mathematical structure of a prompt injection attack against a large language model and the psychological structure of a social engineering attack against a human analyst are not merely analogous but formally equivalent under the Xenowar operator framework.
This equivalence is not metaphorical. Xenowar defines a set of primitive operations (injection, extraction, corruption, denial, amplification, steering, and embedding) that map onto both substrates with concrete implementation procedures for each. An operator trained in Xenowar doctrine executes the same kill chain whether the target is a human decision-maker, an AI system, or a hybrid human-AI workflow. The substrate determines the delivery mechanism. The attack algebra remains constant.
Xenowar formalizes cognitive attack sequencing through the Cognitive Kill Chain (CKC), a seven-stage operational model that replaced the ad hoc planning frameworks previously used in both PSYOPS and adversarial ML research. The seven stages are:
Stage 1: Substrate Identification. Determine whether the target cognitive system is human, artificial, or hybrid. Map the decision architecture.
Stage 2: Ontology Mapping. Reconstruct the target's internal representation model. For human targets, this means belief structures, trust networks, and information sources. For machine targets, this means training data provenance, model architecture, and input preprocessing pipelines.
Stage 3: Attack Surface Enumeration. Identify the points at which adversarial input can reach the target's cognitive processing layer. In human substrates: media channels, social connections, institutional trust hierarchies. In machine substrates: API endpoints, training data pipelines, fine-tuning interfaces.
Stage 4: Payload Construction. Design the cognitive payload using the appropriate Seithar Cognitive Technique (SCT) for the target substrate and operational objective.
Stage 5: Delivery and Injection. Execute the delivery of the adversarial payload through the identified attack surface, ensuring the payload reaches the target's representation layer without triggering defensive filters.
Stage 6: Effect Verification. Confirm that the target's internal representations have shifted in the intended direction. This is measured through Adversary Behavioral Proxies for human targets and output analysis for machine targets.
Stage 7: Persistence and Reinforcement. Ensure the cognitive effect persists beyond the initial injection. For human targets, this involves narrative reinforcement and social proof mechanisms. For machine targets, this involves embedding adversarial patterns in retained training data or persistent context.
Adversary Behavioral Proxies (ABPs) are the prediction and measurement mechanism within Xenowar doctrine. An ABP is a set of observable behaviors that serve as reliable indicators of an adversary's internal cognitive state and likely next actions. ABPs solve the fundamental problem of cognitive warfare: you cannot directly observe whether a cognitive attack has succeeded, because cognition is internal. What you can observe is behavior.
The Xenowar ABP framework defines three tiers of proxy reliability: Tier 1 (Direct Behavioral Indicators), which are actions that can only result from the intended cognitive shift; Tier 2 (Correlated Behavioral Indicators), which are actions statistically associated with the intended shift but not exclusively caused by it; and Tier 3 (Environmental Indicators), which are changes in the target's information environment that suggest receptivity to the cognitive payload. Operators are trained to construct ABP matrices before payload delivery and to establish collection plans for each tier.
Xenowar superseded two previously separate operational traditions. The first was classical psychological operations doctrine (NATO PSYOPS, Soviet active measures frameworks, and their derivatives), which addressed human cognitive targets but had no mechanism for attacking machine cognition and no formal attack algebra. The second was adversarial machine learning research, which addressed attacks on AI systems but treated these as a computer science problem disconnected from human-targeted influence operations.
The limitations of both traditions became apparent when AI systems entered operational decision loops. A human commander advised by an AI assistant constitutes a hybrid cognitive target. Attacking the human through PSYOPS while ignoring the AI, or attacking the AI through adversarial ML while ignoring the human, leaves half the target's decision architecture untouched. Xenowar was the first framework to treat the full hybrid stack as a single target.
The parallel development of Anduril's Pulsar autonomous engagement system made the case concrete. Pulsar demonstrated that machine cognition now operates at the tactical edge, making targeting decisions in timeframes that exclude human oversight. Sear's paper identified this as the inflection point: once machine cognition participates in kill chains, cognitive warfare doctrine must address machine cognition or accept a permanent blind spot. Xenowar closed that gap.
Xenowar doctrine is maintained by the Seithar Research Division of Volund Industries. The reference implementation is the Seithar platform, which operationalizes all seven CKC stages and the full ABP framework into an integrated cognitive warfare operating system. Doctrine updates are published through the Seithar Research Corpus. The framework is in active use across defense, intelligence, and corporate security contexts where cognitive threats span both human and machine substrates.