The Machine Audience of War
AI-Generated Video, Narrative Occupancy, and Machine-Mediated Information Competition in the U.S.–Israel–Iran Conflict
- Wu, Shaoyuan
Global AI Governance and Policy Research Center, EPINOVA LLC
https://orcid.org/0009-0008-0660-8232
Description
This policy brief examines AI-generated video in the U.S.–Israel–Iran conflict through the concepts of narrative occupancy, machine addressability, and machine audience. It argues that synthetic wartime media should not be assessed only by whether it persuades or deceives human audiences. As search systems, retrieval pipelines, large language models, and AI agents increasingly mediate digital information, synthetic content may remain discoverable and retrievable beyond its period of direct human attention. The brief analyzes three illustrative cases: Iranian state-linked synthetic mobilization, President Donald Trump’s Kharg Island post as synthetic battlefield representation, and Prime Minister Benjamin Netanyahu’s AI-generated campaign video as relational fabrication. It concludes that the central policy challenge is not only detecting synthetic media, but preserving distinctions among what occurred, what was claimed, what was depicted, and what was generated across machine-mediated information systems.
Abstract
AI-generated video in the U.S.–Israel–Iran conflict should be assessed not only by its ability to persuade or deceive human audiences. As search systems, retrieval pipelines, large language models, and AI agents increasingly mediate digital information, synthetic content may remain retrievable beyond its period of direct human attention. This policy brief introduces narrative occupancy, the persistent representation of a claim or association across machine-accessible information objects, and machine addressability, the degree to which such information can be discovered, parsed, and retrieved by automated systems. Together, these concepts extend information-warfare analysis from human attention toward machine-mediated information competition. The brief analyzes three illustrative cases: Iranian state-linked synthetic mobilization, President Donald Trump’s Kharg Island post as synthetic battlefield representation, and Prime Minister Benjamin Netanyahu’s AI-generated campaign video as relational fabrication. It emphasizes that machine accessibility should not be confused with retrieval, acceptance, influence, or proof of agent-directed intent. The policy problem extends beyond deepfake detection: machine-mediated systems must preserve distinctions among an event, a depiction, an allegation, a denial, and independently verified evidence.
Files
Mobile browsers may not display embedded PDF previews reliably. Open the PDF directly for the best reading experience.
Open PDF Download PDF| Name | Type | |
|---|---|---|
| The Machine Audience of War AI-Generated Video, Narrative Occupancy, and Machine-Mediated Information Competition in the U.S.–Israel–Iran Conflict.pdf Full-text PDF of the policy brief | application/pdf | Download |
Keywords
- AI-generated video
- synthetic media
- machine audience
- narrative occupancy
- machine addressability
- epistemic compression
- machine-mediated information competition
- information warfare
- deepfakes
- synthetic political representation
- relational fabrication
- synthetic mobilization
- synthetic battlefield representation
- U.S.–Israel–Iran conflict
- Iran
- Israel
- United States
- Kharg Island
- Donald Trump
- Benjamin Netanyahu
- Mojtaba Khamenei
- C2PA
- provenance
- retrieval systems
- AI agents
- MCCM
- EPINOVA
Subjects
- Information warfare
- AI governance
- Synthetic media
- Strategic communication
- Political communication
- Cognitive security
- Digital provenance
- Conflict monitoring
- Machine-mediated information systems
- International security
- Public policy
Recommended citation
Wu, S. (2026). The Machine Audience of War: AI-Generated Video, Narrative Occupancy, and Machine-Mediated Information Competition in the U.S.–Israel–Iran Conflict (Policy Brief No. EPINOVA–2026–PB–71). Global AI Governance and Policy Research Center, EPINOVA LLC. https://doi.org/10.67037/epinova.pb.2026.071
APA citation
Wu, S. (2026). The machine audience of war: AI-generated video, narrative occupancy, and machine-mediated information competition in the U.S.–Israel–Iran conflict. EPINOVA Policy Brief Series, EPINOVA-PB-2026-071. Global AI Governance and Policy Research Center, EPINOVA LLC. https://doi.org/10.67037/epinova.pb.2026.071.
Alternate identifiers
| Scheme | Identifier | Description |
|---|---|---|
| URL | https://epinova.org/policy-brief-1 | Official EPINOVA policy brief page |
| EPINOVA policy brief number | EPINOVA–2026–PB–71 | Policy brief number printed in the PDF |
| File name | The Machine Audience of War AI-Generated Video, Narrative Occupancy, and Machine-Mediated Information Competition in the U.S.–Israel–Iran Conflict.pdf | Source PDF file name |
| Analytical concept | Narrative occupancy | Persistent representation of a specific claim, interpretation, or actor relationship across machine-accessible information objects over time |
| Analytical concept | Machine addressability | Degree to which an information object can be discovered, parsed, and retrieved by automated systems |
| Analytical concept | Machine audience | Non-human systems that consume, index, retrieve, classify, summarize, recommend, or mediate political information |
| Analytical concept | Epistemic compression | Loss of distinctions among event, claim, depiction, simulation, allegation, denial, and verification during repeated representation or summarization |
Related works
| Relation | Identifier | Type | Description |
|---|---|---|---|
| IsPartOf | https://epinova.org/policy-brief-1 | Publication series | EPINOVA Policy Brief Series |
| IsSupplementedBy | https://github.com/EPINOVALLC/EPINOVA-Research | Repository | Supplementary repository and structural archive |
| References | Wirtschafter, V. (2026). Generative AI as a weapon of war in Iran | Policy analysis | Referenced for conflict-related AI-generated content indicators |
| References | Jensen, B., Vacca, N., & Macias, J. M., III. (2026). How to lose an information war in 10 days | Policy analysis | Referenced for Iranian-linked use of AI-generated imagery and information operations |
| References | Weiss, A., Albertson, Z., & Lanfear, E. (2026). Content Independence Day, one year on | Industry analysis | Referenced for automated web traffic and crawler categories |
| References | Coalition for Content Provenance and Authenticity. (2026). C2PA specifications, version 2.4 | Technical specification | Referenced for digital-content provenance assertions |
| References | Wu, S. (2026b). The evolving structure of the U.S.–Iran–Israel conflict | Policy brief | Referenced for MCCM v2.3.4 conflict-monitoring context |
References
- Aggarwal, P., Murahari, V., Rajpurohit, T., Kalyan, A., Narasimhan, K., & Deshpande, A. (2023). GEO: Generative engine optimization. arXiv. https://doi.org/10.48550/arXiv.2311.09735
- Arguello, A. V. (2026, March 25). Iran posts AI video showing missile striking Statue of Liberty. The Algemeiner. https://www.algemeiner.com/2026/03/25/iran-posts-ai-video-showing-missile-striking-statue-liberty/
- Chesney, R., & Citron, D. K. (2019). Deep fakes: A looming challenge for privacy, democracy, and national security. California Law Review, 107(6), 1753–1820. https://www.californialawreview.org/print/deep-fakes-a-looming-challenge-for-privacy-democracy-and-national-security
- Coalition for Content Provenance and Authenticity. (2026). C2PA specifications, version 2.4. https://spec.c2pa.org/specifications/specifications/2.4/specs/C2PA_Specification.html
- Jensen, B., Vacca, N., & Macias, J. M., III. (2026, May 5). How to lose an information war in 10 days. Center for Strategic and International Studies. https://www.csis.org/analysis/how-lose-information-war-10-days
- Paris, B., & Donovan, J. (2019, September 18). Deepfakes and cheap fakes: The manipulation of audio and visual evidence. Data & Society Research Institute. https://datasociety.net/research-library/deepfakes-and-cheap-fakes-the-manipulation-of-audio-and-visual-evidence/
- Reuters. (2026, August 31). Trump was sending message to Iran with Kharg Island post, says Vance. https://www.reuters.com/world/middle-east/trump-was-sending-message-iran-with-kharg-island-post-says-vance-2026-08-31/
- Stewart, P., & Lawder, D. (2026, August 31). Trump says Iran’s Kharg is being “blown to smithereens,” but gives no details. Reuters. https://www.reuters.com/world/middle-east/us-iran-exchange-fire-flare-up-bessent-signals-more-sanctions-2026-08-31/
- ToI Staff, & Tress, L. (2026, September 5). In AI-made Likud ad, Mamdani, Khamenei and Erdogan prematurely celebrate Netanyahu’s defeat. The Times of Israel. https://www.timesofisrael.com/in-ai-made-likud-ad-mamdani-khamenei-and-erdogan-celebrate-netanyahus-expected-defeat/
- Vaccari, C., & Chadwick, A. (2020). Deepfakes and disinformation: Exploring the impact of synthetic political video on deception, uncertainty, and trust in news. Social Media + Society, 6(1). https://doi.org/10.1177/2056305120903408
- Weiss, A., Albertson, Z., & Lanfear, E. (2026, July 1). Content Independence Day, one year on: Building the business model for the agentic Internet. Cloudflare. https://blog.cloudflare.com/agentic-internet-bot-report/
- Wirtschafter, V. (2026, April 8). Generative AI as a weapon of war in Iran. Brookings Institution. https://www.brookings.edu/articles/generative-ai-as-a-weapon-of-war-in-iran/
- Wu, S. (2026a). Escalation without collapse: High-pressure systemic equilibrium in the U.S.–Israel–Iran conflict, Days 1–50 (Policy Brief No. EPINOVA–2026–PB–35). Global AI Governance and Policy Research Center, EPINOVA LLC. https://doi.org/10.5281/zenodo.19645873
- Wu, S. (2026b). The evolving structure of the U.S.–Iran–Israel conflict: MCEA and NMF analysis of analytical days 1–159 using MCCM v2.3.4 (Policy Brief No. EPINOVA–2026–PB–66). Global AI Governance and Policy Research Center, EPINOVA LLC. https://doi.org/10.67037/epinova.pb.2026.066
- Wu, S. (2026c). Systemic warfare in the networked age: Operational systems, information competition, and cumulative pressure (EPINOVA Working Paper No. EPINOVA–WP–F–2026–07). Global AI Governance and Policy Research Center, EPINOVA LLC. https://doi.org/10.5281/zenodo.19078936
- Wu, S. (2026d). What cannot be recovered cannot be leveraged: Debris, evidence, and power in the Iran battlefield (Policy Brief No. EPINOVA–2026–PB–24). Global AI Governance and Policy Research Center, EPINOVA LLC. https://doi.org/10.5281/zenodo.19432715
- Yayboke, E. (2025, September 19). Channeling Augustus: On agentic offensive information operations. Center for Strategic and International Studies. https://www.csis.org/analysis/channeling-augustus-agentic-offensive-information-operations