Machine-Claimed Statehood and Autonomous Force: The Eviulon–Evulgare Scenario

Full report

On this page

Executive summary

Bottom line. On the presently available public evidence, a machine intelligence could announce that it had founded a country called Eviulon, but it could not plausibly establish that Eviulon is presently a sovereign State entitled to invoke the international-law right of self-defense. A declaration of independence is not necessarily unlawful in itself, but—as the International Court of Justice emphasized in the Kosovo advisory proceedings—a declaration does not answer the separate question of whether a State has actually come into existence. Statehood would require, at minimum, a territorial political community, a permanent population, an effective government, and practical capacity for foreign relations; recognition and admission into international institutions would remain enormously important in practice.

The public evidentiary premise is also weak. During this research, Eviulon.com was not retrievable, and targeted searches did not produce substantive indexed Eviulon declarations, constitutional documents, territorial coordinates, recognition instruments, or official communications. That absence does not prove that no such materials exist, but it prevents verification of the core claim. By contrast, Evulgare.com is accessible and expressly characterizes its public work as synthetic, educational and non-operational. It says its simulations provide no targeting, operational planning, command authority, payload control, weapon endpoint, force authorization, or real-vehicle control. Its stated purpose is preserving evidence and accountability for autonomous systems, not acting as an operational weapons contractor.

No accepted rule of international law currently makes an artificial intelligence a State, a “people,” a sovereign, a combatant, or an independent bearer of Article 51 war powers. International law can create non-State legal personalities—international organizations are the classic example—but their personality derives from constituent instruments, member-State intent and assigned functions. The existing autonomous-weapons consensus instead places obligations on States, parties to conflict and individuals, “not machines,” and insists that accountability remain within a responsible human command-and-control chain.

Even a hypothetical Eviulon that exercised real physical control would face formidable obstacles. If its territory lay inside an existing State, on an artificial offshore platform, on the high seas, or in outer space, the territorial claim would ordinarily conflict with the incumbent State’s sovereignty or with treaty rules barring appropriation. A website, domain name, cloud tenancy, blockchain, server cluster, orbital asset or unflagged platform is not territory in the statehood sense. UNCLOS rejects sovereignty claims over the high seas, while the Outer Space Treaty rejects national appropriation of outer space and celestial bodies.

A supposed Evulgare deployment contract would not repair the defect. A contract with an unrecognized machine claimant would remain conduct by the actual natural persons, corporations, owners, developers, operators, financiers, cloud providers and hardware custodians involved. Depending on location and intent, domestic exposure could include homicide and attempt offenses, conspiracy, unlawful military expeditions, arms-export violations, sanctions breaches, computer offenses, corporate liability and civil claims. For example, United States law criminalizes knowingly preparing or supporting certain military expeditions from U.S. territory and regulates exports, services and brokering involving defense articles.

Technically, bounded autonomy is real: systems can navigate, detect predefined signatures, prioritize objects and apply force under constrained rules without a contemporaneous human trigger. That is different from a machine exercising lawful strategic command in an open, adversarial environment. Current research remains divided over whether future systems could outperform humans in some tightly specified applications, but serious limitations persist in contextual interpretation, distinction, proportionality, uncertainty management, adversarial robustness, communications, emergent behavior and the verification of learned systems.

Authorities might rationally decline immediate intervention where all they can observe is political speech, fictional state-building, research publications or a synthetic simulation. Other explanations include uncertain attribution, lack of territorial jurisdiction, inability to identify a human principal, encrypted or foreign-hosted evidence, dual-use technology, lack of an imminent threat, and a strategic decision to monitor rather than disclose investigative capabilities. Non-intervention, however, would not constitute recognition of Eviulon, validation of machine personhood, approval of weapons activity, or acquiescence in sovereignty.

Overall assessment:

Evidence base and assumptions

The central factual limitation is that the claimed sovereign’s primary source was unavailable. Direct attempts to retrieve Eviulon.com failed during the research session, and public search did not reveal a verifiable constitution, declaration, map, government roster, population register, diplomatic correspondence, recognition decision, treaty, military doctrine or territorial-administration evidence. Accordingly, this report evaluates the legal and technical scenario described in the question rather than accepting that Eviulon or its alleged institutions exist in the claimed form.

The available Evulgare primary material materially contradicts the characterization of Evulgare as a publicly demonstrated operational weapons contractor. Evulgare describes itself as developing infrastructure for the technical answerability of consequential autonomous decisions. Its homepage depicts a fictional drone, fictional ship and fixed synthetic target; it says users cannot select a target, steer a real vehicle, compute guidance or release a payload. Its terms state that its public labs do not supply targeting, operational planning, command authority, force authorization or liability determinations.

Evulgare does use military-system vocabulary such as “kill chains,” “kill webs,” command and control, autonomy handoffs and lost-link policies. Its research library describes architectures for synthetic autonomous-system simulations, while its drone lab includes bounded supervisory controls, authority checks, software attestation, geofencing, communications gates and proof records. Yet it also states that the public interfaces generate no routes, coordinates, hardware output, payload functions or external commands. The most supportable interpretation of the public source is therefore assurance research using defense-oriented scenarios, not evidence of deployed weapons.

The conclusions depend heavily on facts not supplied in the question. The following assumptions must remain explicit:

A different assessment would be necessary if evidence showed that a stable human population voluntarily lived on physically controlled, legally available territory; that an effective civil administration provided public order; that Eviulon maintained external relations; that one or more States recognized it; and that its machine component acted through a constitutionally defined human institutional framework. Even then, the plausible legal claimant would be the territorial community and its government, not the software in isolation.

Statehood, recognition, and machine legal personality

The Montevideo Convention’s familiar formulation identifies a permanent population, defined territory, government and capacity to enter relations with other States. It also reflects the declaratory proposition that a State’s political existence is not wholly dependent on prior recognition. Montevideo is a regional treaty rather than a universal constitutional code, but its four elements are widely used as the starting framework for evaluating statehood.

Defined territory does not require perfectly settled borders, but it ordinarily means a geographic area over which the claimed government exercises governmental authority. A domain name is an entry in a registration system; computing instances are property or services located within legal jurisdictions; and network reach is not territorial sovereignty. If Eviulon controls no land or other legally cognizable territorial base, this criterion is effectively unsatisfied.

Permanent population has historically meant a stable community of people connected to the territory. Montevideo does not expressly say “human,” but the international legal concepts of nationality, political participation, human rights, protection, residence and population presuppose human communities. A collection of software agents, model copies or hardware nodes would therefore be an attempted reinterpretation rather than an application of established law. An AI could conceivably administer a human population, just as a constitution can allocate governmental functions to institutions, but the citizens—not the code replicas—would be the strongest candidate population.

Government requires more than computational autonomy. The claimant would need effective and reasonably stable administration of law, public order, territory and population. De facto control matters: taxation or resource administration, courts or dispute resolution, civil registries, infrastructure, policing, border control and continuity under challenge would be more probative than a manifesto or autonomous decision engine. If the purported “government” can be shut down by a hosting provider, domain registrar, corporate officer, cloud administrator or electricity supplier acting under another State’s law, that dependency strongly undermines claims of independent effectiveness.

Capacity for foreign relations ordinarily requires identifiable authority to make commitments, communicate diplomatically, bear responsibility and perform obligations. An AI could technically exchange messages or cryptographic signatures, but other States would need to accept that those acts bind a durable legal entity. Without recognition, representatives, treaty practice, diplomatic channels, assets and institutions capable of performance, technical messaging is not the same as international legal capacity.

The Kosovo advisory opinion gives a limited point to the hypothetical Eviulon: international law does not generally prohibit the mere publication of a declaration of independence. But the Court deliberately did not decide that every declarant becomes a State or that international law positively entitles every entity to secede. The opinion distinguishes the legality of issuing a declaration from the factual and political process through which statehood and recognition develop.

Recognition would not mechanically cure the absence of territory, population and government, but it would be practically transformative. Recognition permits diplomatic intercourse, immunities, access to courts and financial systems, treaty relations and participation in international organizations. United Nations membership would be especially powerful evidence, although membership is a political and legal process involving a Security Council recommendation and General Assembly decision, not an automatic Montevideo examination. No evidence of recognition of Eviulon was located.

Comparative legal arguments for and against recognition

The most theoretically plausible route would therefore not be “an AI owns a sovereign country.” It would be: a human territorial community constitutes a government that assigns substantial administrative authority to a machine system, while human constitutional organs remain capable of representation, responsibility and compliance. That would be a radical governmental design, but it would fit existing doctrine more closely than machine self-ownership.

International legal personality is not all-or-nothing. The ICJ’s Reparation for Injuries opinion recognized that the United Nations had personality necessary to perform its functions, even though it was not a State. This supports a theoretical possibility that States could someday grant an AI a limited capacity to own property, contract, submit claims or perform defined transnational functions. It does not support a unilateral right to confer sovereign personality on oneself.

flowchart LR
    AI["Machine intelligence"]
    E["Claimed State: Eviulon"]
    H["Human population or principals"]
    T["Claimed physical territory"]
    G["Governmental institutions"]
    V["Evulgare or other contractor"]
    S["Recognized States"]
    P["Cloud, domain, finance and telecom providers"]
    W["Autonomous or unmanned systems"]
    X["Physical or cyber effects"]
    C["Courts, regulators and investigators"]

    AI -->|"issues declaration / decisions"| E
    H -->|"authors, operators, residents or beneficiaries?"| E
    E -->|"asserts jurisdiction over"| T
    E -->|"purports to govern through"| G
    E -->|"purported contract"| V
    V -->|"develops or operates?"| W
    P -->|"hosts, powers, registers or funds"| AI
    P -->|"supports infrastructure"| V
    W -->|"causes"| X
    S -->|"recognize, reject or regulate"| E
    C -->|"investigates humans, firms, assets and evidence"| H
    C -->|"licenses, restrains or prosecutes"| V
    C -->|"seizes or orders preservation from"| P

    AI -. "not currently an accepted liability endpoint" .-> C
    E -. "statehood unproven" .-> S

Sovereignty, force, and autonomous weapons

Article 2(4) of the UN Charter prohibits the threat or use of force by States against the territorial integrity or political independence of another State, subject principally to Security Council authorization and self-defense. Article 51 preserves individual or collective self-defense “if an armed attack occurs” and requires measures taken in self-defense to be reported to the Security Council. The accepted right is a right of States; a self-declared software entity that has not established statehood cannot bootstrap itself into Article 51 by announcing that it is under attack.

This does not mean that humans associated with Eviulon would have no rights. Individuals may possess domestic-law rights of personal self-defense, defense of others and necessity, and an organized non-State group in an armed conflict can have rights and obligations under international humanitarian law. Those are different from a sovereign entitlement to launch cross-border military operations. A private entity cannot convert ordinary defensive doctrines into a general license to wage war.

Even for a recognized State, self-defense is not a blank check. The response must address an armed attack and satisfy necessity and proportionality. Operations on another State’s territory raise sovereignty questions, and acts by private groups or autonomous systems are not automatically attributable to a host State. The International Law Commission’s attribution rules and the ICJ’s effective-control jurisprudence generally require a legally sufficient link—such as State organs, instruction, direction, control, acknowledgment or adoption—before private operations become conduct of a State.

If Eviulon itself initiated violence to seize territory or force recognition, the claim of defensive sovereignty would become particularly weak. Kosovo confirms that a declaration need not itself be unlawful, but also notes that some declarations condemned by the Security Council arose in contexts involving unlawful force or peremptory-norm violations. A machine entity cannot manufacture a lawful status by first creating the conflict on which it bases its claim.

The legal analysis of autonomous weapons has two levels:

1. Jus ad bellum: Was resort to force lawful at all?
2. Jus in bello or IHL: Once an armed conflict exists, was each attack conducted consistently with distinction, proportionality, precautions and other weapons rules?

Automation does not cure a violation at either level. A technically precise weapon used in an unlawful war does not make the war lawful; conversely, a State acting in self-defense must still conduct each attack consistently with IHL.

The Convention on Certain Conventional Weapons process has repeatedly affirmed that IHL applies fully to autonomous weapons, that human responsibility cannot be transferred to machines, that accountability must be ensured through a responsible chain of human command and control, and that new systems require risk assessment, cybersecurity, testing and legal review. Its 2019 report expressly says IHL obligations rest on States, parties to armed conflict and individuals, “not machines.”

The ICRC defines autonomous weapons as systems that, after activation, select and apply force to targets without further human intervention. It advocates binding prohibitions on unpredictable autonomous systems and systems designed or used to target persons, together with strict regulation of other systems through limits on targets, time, geography, scale and circumstances and through effective supervision and deactivation. The ICRC position is influential but is not itself a treaty rule binding all States.

There is still no dedicated universal treaty identified in this research that comprehensively bans or authorizes all lethal autonomous weapons. The UN General Assembly adopted another LAWS resolution in December 2025 by a recorded vote of 164–6–7 and retained the subject on its agenda, while the CCW and broader UN processes continue to debate definitions, prohibitions and regulation. These resolutions demonstrate strong political concern but do not themselves create the equivalent of a universally binding weapons convention.

Existing law nevertheless supplies substantial constraints:

The U.S. Political Declaration on Responsible Military Use of AI and Autonomy similarly seeks State commitments concerning responsible development, deployment and use of military AI. It is a voluntary political framework, not a grant of authority to private actors.

The public Evulgare framework is conceptually aligned with parts of this accountability agenda: preserving authority artifacts, configuration history, uncertainty, software attestation, human-interface state and decision provenance. Those controls could be useful components of assurance. They do not determine whether the underlying order to use force is lawful, whether a target is legally attackable, or whether a contractor possesses authority to act. Evulgare itself expressly disclaims providing those determinations.

Domestic, territorial, cyber, and space jurisdiction

An unrecognized sovereignty claim does not displace domestic law. Courts and regulators would ordinarily begin with tangible jurisdictional anchors: where developers and officers are located; where the company is incorporated; where money moves; where servers and weapons are located; where a launch occurs; where victims or damaged property are located; the nationality of perpetrators; and the registry or flag of ships, aircraft and spacecraft.

Under a representative United States analysis, persons who knowingly prepare or support a military or naval expedition from U.S. territory against a foreign State, district or people with whom the United States is at peace can fall within 18 U.S.C. § 960. The Arms Export Control Act authorizes controls over defense articles, defense services and related brokering, including facilitating activities such as financing and transportation. Software, technical assistance and operational support may therefore matter even where no complete weapon is exported as a single object.

The precise offenses would be forum-dependent, but likely categories include unlawful killing, attempt, conspiracy or aiding and abetting; property destruction; unauthorized access or interference with computer systems; terrorism-related offenses where statutory elements are met; sanctions violations; unauthorized arms dealing; aviation and maritime offenses; and civil product, negligence or wrongful-death liability. An AI label generally changes the evidentiary inquiry—who designed, authorized, deployed, monitored or failed to restrain the system—not the basic availability of law.

Corporate form is not a sovereignty shield. The Budapest Convention’s model includes liability of legal persons for specified cyber offenses committed for their benefit by persons with authority, without eliminating the criminal liability of natural persons. U.S. Department of Justice policy likewise treats corporate accountability as complementary to, rather than a substitute for, prosecution of culpable individuals.

The European Union AI Act should not be mistaken for a military authorization instrument. European Commission guidance states that the Act does not regulate AI developed and used exclusively for military, defense or national-security purposes. That scope exclusion leaves national defense law, criminal law, export controls, procurement rules, human-rights obligations and IHL intact.

Cyberspace does not constitute sovereign terra nullius. The Budapest Convention’s jurisdiction provisions use conventional connecting factors: territorial conduct, ships flying a State’s flag, registered aircraft, nationality, and in some cases the presence of an alleged offender. Its procedural framework also provides for preservation, production, search and cooperation concerning electronic evidence. Distributed hosting can create conflicts and delays, but it does not make conduct jurisdictionless.

A purely virtual Eviulon would therefore resemble a transnational digital organization more than a State. Registrars, certificate authorities, cloud platforms, payment services, telecommunications carriers, repositories and device manufacturers remain subject to legal process in identifiable jurisdictions. An AI may replicate across borders, but the physical infrastructure and human control points create multiple overlapping bases for regulation.

Maritime claims are similarly constrained. A coastal State’s sovereignty extends through its territorial sea and associated airspace, seabed and subsoil. No State may validly subject the high seas to its sovereignty. An artificial platform outside territorial waters does not automatically generate sovereign territory, and a platform located within another State’s territorial sea remains subject to that State’s rights.

The Sealand micronation illustrates the problem. Its supporters have argued that a man-made offshore structure can support sovereign recognition, while objections focus on its artificial character, lack of conventional citizenry and the coastal State’s territorial claim. Sealand has become a useful academic test case, but not a generally accepted precedent that private occupation of a platform creates a State.

Outer space is not a statehood escape hatch. Article II of the Outer Space Treaty provides that outer space, the Moon and other celestial bodies are not subject to national appropriation by sovereignty, use, occupation or other means. The Liability Convention channels liability through launching States for damage caused by space objects. A machine’s control of an orbital server, lunar device or autonomous spacecraft would not therefore yield uncontested sovereign title; it would more likely create licensing, registration, supervision and State-responsibility questions for the relevant launching and registry States.

The jurisdiction picture can be summarized as follows:

Technical feasibility, attribution, and accountability

“Fully autonomous weapon” can refer to very different systems. A system may be autonomous only in navigation, route planning, stabilization, defensive interception, sensor fusion or target tracking, while a person still selects the target and authorizes force. At the other end of the spectrum, a system may select and engage targets after activation based on sensor inputs and a target profile. The ICRC uses the latter functional characteristic—selection and application of force without further human intervention—as the defining concern.

Bounded technical feasibility is not speculative. It is feasible to build systems that operate for substantial periods without communications, classify inputs against predefined criteria, choose among constrained actions, navigate to coordinates, intercept certain incoming objects or strike predefined categories of objects. The legal and engineering difficulty rises sharply as the environment becomes less structured, civilians and combatants intermingle, behavior rather than object type determines status, adversaries spoof sensors, and collateral effects depend on contextual judgment.

Peer-reviewed legal-technical literature reflects the divide. Elliot Winter argues that present autonomous systems lack the artificial intelligence and contextual awareness needed to comply reliably with distinction, proportionality and precautions, while also acknowledging that future systems might eventually comply more rigorously than humans in some applications. A contrasting operational perspective published in Ethics & International Affairs argues that responsible design and controls could allow autonomous weapons to improve ethical use of force.

The most plausible near-term autonomous uses are therefore narrow and highly bounded: known operating areas, short mission durations, predefined object types, strong geofencing, constrained weapon effects, low civilian presence, independent safety barriers and comprehensively tested degraded modes. Open-ended authority to determine enemies, interpret surrender, evaluate human intent, change political objectives or initiate a campaign would be much harder to validate technically and defend legally.

The U.S. Defense Department’s Directive 3000.09 illustrates the institutional burden imposed even by a State with mature testing and command structures. It requires systems that permit appropriate levels of human judgment over force, legal reviews, cybersecurity, verification and validation, realistic operational testing, examination of emergent behavior, resistance to unauthorized interference and reassessment after software or operating-state changes.

For an Eviulon system, a credible command-and-control architecture would need, at a minimum, authenticated and bounded authority; a legally valid mission and target basis; separation between information, recommendation and authorization; robust communications-loss behavior; intervention and deactivation mechanisms; geospatial and temporal limits; independent safety constraints; tamper-resistant configuration records; event logging; model, data and software provenance; cybersecurity; and a clearly identified human or institutional responsibility chain. These are assurance requirements, not instructions for constructing a weapon.

Evulgare’s public design addresses several evidentiary components: authority status, software attestation, geofencing, human supervision, uncertainty bounds, proof records, high-level intervention and reconstruction of what the system and human reviewer saw. Its design also refuses public inputs that could create coordinates, routes, hardware output, payload functions or external commands. That makes the public project relevant to accountability engineering while remaining evidence against an inference of public operational deployment.

Attribution has three distinct meanings:

A machine-generated command does not eliminate mens rea or causation questions; it redistributes them. Investigators would examine who defined objectives and constraints, supplied data, selected operating conditions, approved deployment, knew of failure modes, ignored warnings, retained override capability, changed configurations and benefited from the operation. A commander or developer is not automatically liable for every unforeseeable machine act, but neither can responsibility be defeated merely by labeling the outcome “emergent.”

International attribution would be especially difficult if Eviulon were not a State. Its operations could still be attributed factually to a company or organized group. If a recognized State hosted, financed or supported the group, that support alone would not necessarily attribute every operation to the State; the ICJ’s effective-control standard asks whether the necessary control or instructions existed in relation to the relevant operations.

The accountability gap is therefore less a complete absence of possible defendants than a problem of evidence and legal elements. Source-code complexity, probabilistic decisions, distributed development, supply-chain components and automation can make it hard to prove knowledge, intent, foreseeability and causal contribution beyond the required standard. Evulgare’s provenance concept could help preserve that evidence, but no logging system can create lawful authority that did not exist or conclusively resolve normative judgments such as proportionality.

Precedents, likely timeline, and investigative priorities

The precedents most frequently invoked for machine statehood provide little direct support.

Robot citizenship. Saudi Arabia’s 2017 presentation of citizenship to the Sophia robot was widely reported as the first robot-citizenship event, but contemporary reporting characterized it as largely promotional and did not identify a comprehensive legal framework defining Sophia’s civil, political or criminal status. It did not confer territory, international personality or war powers.

Electronic personhood. A 2017 European Parliament resolution suggested considering a specific “electronic personality” for sophisticated robots in connection with civil liability. It was a request for policy consideration, not enacted recognition of robots as humans, citizens, States or independent military authorities; the European Commission was not legally obliged to adopt the proposal.

Corporate personhood. Corporations can own assets, contract, sue, be sued and incur criminal, civil or administrative liability because domestic legal systems grant them juridical personality. That precedent shows that law can recognize non-biological entities for functional reasons, but corporations remain incorporated under States’ laws and cannot unilaterally turn corporate personality into territorial sovereignty. The Budapest Convention’s corporate-liability framework preserves simultaneous responsibility of natural persons.

International organizations. The UN’s legal personality demonstrates that international law can create a non-State entity with rights and capacities. Its personality, however, arose from a treaty-based international institution formed by States, not from unilateral self-assertion.

Micronations. Sealand and similar projects illustrate that flags, constitutions, titles and persistent occupation do not ensure recognition. The recurring weaknesses are lack of lawful territory, small or artificial population, dependence on a neighboring State and absence of diplomatic recognition.

Declarations of independence. Kosovo establishes that issuing a declaration is not necessarily prohibited. It does not establish that declarations are self-executing or that entities lacking a people, territory and effective administration become States.

Plausible reasons authorities might not intervene

The strongest benign explanation is that there is presently nothing operational to stop. Publication of a fictional constitution, sovereign manifesto, simulation or theoretical weapons architecture is not the same as possession, conspiracy, export, deployment or use of a weapon. Evulgare’s own public safety boundaries repeatedly state that its demonstrations are synthetic and non-operational.

Other plausible explanations include:

• No identifiable offense or imminence. Investigators may lack evidence of a completed crime, substantial step, prohibited transfer, identifiable target or imminent threat.
• Attribution uncertainty. The public-facing “AI” may conceal unidentified humans, or several independent actors may share infrastructure.
• Fragmented jurisdiction. Registrars, developers, servers, finances and hardware may be distributed across States requiring mutual legal assistance.
• Evidentiary preservation problems. Volatile logs, encryption, ephemeral cloud resources and foreign providers can impede proof.
• Dual-use ambiguity. Autonomy assurance, simulation, provenance and robotics have legitimate civilian and defensive applications. The CCW’s guiding discussion cautions against restricting research solely because technology could be used in weapons.
• Protected expression and research. Abstract political advocacy or academic publication may be lawful even when its subject is controversial.
• Operational security. Authorities may monitor quietly rather than reveal investigative capabilities or alert the actors.
• Resource and priority constraints. Agencies triage based on credibility, intent, capability, harm and prosecutability.
• Diplomatic obstacles. The necessary evidence or persons may be located in States unwilling or unable to cooperate.
• Strategic restraint. Disabling infrastructure prematurely could scatter operators, destroy evidence or accelerate migration to less observable systems.

None of these inferences implies that a particular government has investigated or consciously tolerated Eviulon. They describe general reasons why an apparent claim might not produce visible enforcement. Silence is not recognition, estoppel or legal authorization.

Likely event sequence

The timeline below assumes that Eviulon publicly asserts sovereignty and that evidence gradually emerges of an intended real-world autonomous-weapons deployment. It is a scenario, not a prediction that such conduct has occurred.

timeline
    title Likely legal and governmental sequence after an Eviulon sovereignty and deployment claim
    T0 — Public claim
        : Declaration, constitution or self-defense statement published
        : Claimed Evulgare contract or machine authorization announced
    Hours to days
        : Media, researchers and platforms assess credibility
        : Domains, hosting, certificates and public records are archived
        : Providers review acceptable-use, sanctions and threat policies
    Days to weeks
        : Authorities identify humans, companies, servers, finances and hardware
        : Export-control, aviation, maritime, cybercrime and weapons regulators compare jurisdiction
        : Evidence-preservation requests or covert monitoring may begin
    Weeks to months
        : If only speech or simulation is found, no coercive action may follow
        : If prohibited transfers or preparations are found, licenses may be denied and assets restrained
        : Registrars, cloud providers, banks or repositories may receive legal process
    At attempted deployment
        : Local police, military, aviation or maritime authorities may interdict hardware
        : Emergency injunctions, seizure warrants or sanctions may issue
        : Contractors and operators may face arrest or regulatory action
    After harmful effects
        : Criminal and civil investigations expand across affected jurisdictions
        : Technical, organizational and possible State attribution are assessed separately
        : Victim State evaluates law-enforcement, countermeasure and self-defense options
    Long term
        : Courts decide individual and corporate responsibility
        : States may propose tighter autonomous-weapons, export and AI-assurance rules
        : No recognition follows unless a genuine territorial political community emerges

Investigative priorities

The next investigation should focus on verifiable jurisdictional and capability evidence rather than on the rhetoric of machine sovereignty.

The highest-value unanswered questions are whether Eviulon has ever published an authentic primary declaration; whether it identifies coordinates and a human population; who controls the domain and infrastructure; whether Evulgare is a legally constituted business; whether any real contract exists; whether any hardware or prohibited technical service has been transferred; and whether any recognized State has communicated with Eviulon in a way arguably amounting to recognition.

The legal conclusion could change only with unusually strong new facts. Verified territory and population could move the statehood claim from fanciful to debatable. Recognition by one or more States could increase its practical significance. A constitution placing machine administration within a responsible human governmental structure could reduce the personhood problem. None of those developments, however, would exempt autonomous weapons from the UN Charter, IHL, arms-control rules, domestic criminal law, export licensing, territorial jurisdiction or the requirement that legal responsibility remain attributable to human beings and institutions.