The Fog of War Machines

The global internet functions as a set of invisible veins. Over 1.4 million kilometers of undersea fiber optic cables carry the vast majority of international data traffic. The Pacific Ocean, by virtue of its sheer size and position connecting the Americas to Asia, serves as the grand central station for this physical traffic. This brute fact of geography dictates the architecture of global surveillance. The Five Eyes intelligence alliance, for instance, functions as a geographic agreement as much as a political one. The inclusion of the Australian Signals Directorate (ASD) and New Zealand's Government Communications Security Bureau (GCSB) is a function of their physical location, granting them access to this data flow. This access has a grubby, physical history: Covert hubs disguised in unassuming suburban landscapes, like "Pinecone" in New Jersey or "Room 641A" in San Francisco, serve as landing points where telecom giants collaborate with state agencies to tap the world's data stream. This physical infrastructure also presents a spectacular vulnerability: Near the Taiwanese island of Matsu, undersea cables were severed 27 times in five years. This pattern signals a clear use of asymmetric cyber and information warfare. Why develop an algorithm to decrypt an enemy's data stream when one can just cut the wire?

The competition for cable access feeds a digital Cold War. This conflict unfolds as a matter of techno-diplomacy, a contest between ideological camps fighting to define the global network's operating system. The first camp, the Western alliance anchored by the Five Eyes, propagates the doctrine of an open internet, secure by design. This model advocates for multistakeholder governance and, at least in its public-facing documents, digital human rights. The second camp, the authoritarian-sovereigntist bloc led by China and Russia, champions cyber sovereignty. This doctrine holds a starkly simple premise: A state possesses the absolute right to control its internal internet, free from external interference. This thinking extends directly from China’s domestic comprehensive national security concept, which elevates political security (the survival of the Chinese Communist Party) above all other considerations.

This domestic system, packaged as "Safe China" and built on programs like Skynet and Sharp Eyes, now gets exported via the Digital Silk Road. Tech giants like Huawei and ZTE provide the affordable, state-subsidized surveillance platforms. This war of models happens now in the Pacific. It is no surprise that China has actively pursued regional security pacts with Pacific Island Countries. When that broad diplomatic effort failed, Beijing pivoted. It now pursues persistent, bilateral police training and security assistance by providing tools. For a Pacific nation, choosing a 5G provider or accepting a Safe City grant now constitutes a binding, long-term ideological decision. The result looks like the Splinternet, a global network fracturing along these geopolitical fault lines.

Now, parallel to the high-level state conflict, a second, profit-driven surveillance war rages. For Pacific-rim nations like Aotearoa, Transnational Serious Organised Crime (TSOC) now poses the number one threat to national security. The Pacific, due to its location and maritime routes, functions increasingly as a key transshipment hub for illicit goods. For, the threat has physical dimensions: Criminals employ unmanned narco-subs (yep, submarines) to evade maritime surveillance and move drugs along Pacific supply routes. The scale of this trade, and the demand that fuels it, beggars belief. In Aotearoa, the estimated annual social harm from just methamphetamine, cocaine, and MDMA reached NZD $1.65 billion in 2024. Wastewater analysis in 2024 recorded peak methamphetamine consumption at over 35kg per week, a more than threefold increase from the 10kg per week measured in 2021. And this criminal ecosystem has its own potent cyber dimension. Offshore organized crime groups target New Zealanders with cyber fraud, as it strips NZD $397 million in direct losses from victims in 2023.

This creates a legitimate, existential threat of becoming overrun by organised crime. So, Pacific nations have an immediate, pragmatic need for better surveillance tools. Now, China's "Safe China" model gets marketed as an ideal, off-the-shelf solution. At forums, Chinese firms display intelligent unmanned equipment and police robot-dogs aimed at public security prevention and control. A Pacific nation, desperate to stop the flow of narcotics, may accept these tools to fight crime but it may only later discover it has imported the entire ideological and technical apparatus of cyber sovereignty.

Then, even as states install these vast surveillance systems, the technological arms race turns against them. The advantage shifts. The old model of finding (fusing data from ubiquitous sensors) finds itself countered by a new doctrine of hiding. In the military, analysts now hypothesize that AI-driven so-called fog-of-war machines will orchestrate deception with robotic decoys. These systems will manufacture uncertainty at a scale that overwhelms the finder's sensors. At the same time, the legal architecture for policing this space has collapsed. The concept of territoriality, the bedrock of conventional jurisprudence, appears ill-equipped to govern a borderless domain. International agreements, like the Budapest Convention on Cybercrime, become less effective given the simple fact that major actors like Russia and China refuse to participate.

Into this legal and technical vacuum, human communication flees. It migrates to encrypted platforms like WhatsApp. Frustrated by end-to-end encryption, state actors abandon high-tech surveillance. They revert to physical, human-level coercion. This reality of lived encryption means that in Brazil and India, authorities simply seize a person's phone and use extrajudicial tactics to force them to unlock their private chats. Surveillance becomes a brutal, non-technical act. When users move to truly anonymous spaces, something else happens. A 2025 study comparing the surface web (Reddit) with the dark web forum Dread found that Dread users, despite operating in a space with more toxic language, exhibit a stronger sense of community membership. They form a community-oriented space, express more positive emotions, and use praising rhetoric to build group cohesion.

This social evolution is not without linguistic consequence. Research in emergent communication explains how. When a community operates with limited data exposure (the very definition of an insular, encrypted, or dark web group) it creates a learning bottleneck. This pressure forces the group to generalize, and this generalization drives the formation of a compositional structure. They invent a new, efficient, and private language. This process mirrors the same principles that cause high-frequency words in natural language to become idiosyncratic and irregular.

The entire paradigm of surveillance therefore shifts. Access to the physical cable fails if the message remains encrypted. Physical access to the decrypted message fails if the message uses an emergent, compositional code that the analyst cannot interpret. Interpreting the code fails without understanding the social motivations and praising rhetoric of the community that created it. Organizations attempting to put boots on the Pacific will discover that technical strategy alone falls short. The final layer of intelligence requires human, linguistic, and cultural fluency. For, the old maps have become obsolete. The new strategic terrain appears fully encoded in meaning.