Tracking Patterns of Policy Violations

The subject deliberately pushes or tests organizational policies, rules, or controls to assess tolerance levels, detect oversight gaps, or gain a sense of impunity. While initial actions may appear minor or exploratory, boundary testing serves as a psychological and operational precursor to more serious misconduct.

Characteristics

• Motivated by curiosity, challenge-seeking, or early-stage dissatisfaction.
• Actions often start small: minor policy violations, unauthorized accesses, or circumvention of procedures.
• Rationalizations include beliefs that policies are overly rigid, outdated, or unfair.
• Boundary testing behavior may escalate if it is unchallenged, normalized, or inadvertently rewarded.
• Subjects often seek to gauge the likelihood and severity of consequences before considering larger or riskier actions.
• Testing may be isolated or gradually evolve into opportunism, retaliation, or deliberate harm.

Example Scenario

A subject repeatedly circumvents minor IT security controls (e.g., bypassing content filters, using personal devices against policy) without immediate consequences. Encouraged by the lack of enforcement, the subject later undertakes unauthorized data transfers, rationalizing the behavior based on perceived inefficiencies and low risk of detection.

The subject exploits their technical role to deploy or manipulate automated bots within the organization’s environment—most commonly within collaboration platforms (e.g., Slack, Teams, Discord) or internal operational systems (e.g., Jira, ServiceNow, Helpdesk tooling). These bots are designed to persist beyond the subject’s tenure, leveraging independent service credentials (or other credentials not specifically associated to a user), webhook integrations, or unattended workflows to maintain covert access.

The subject may create new bots under the guise of legitimate productivity enhancements, or hijack existing integrations to expand data access, redirect output, or embed hidden monitoring functionality. Once active, these bots operate continuously, harvesting internal conversations, extracting files, or polling sensitive endpoints—often without triggering standard audit alerts tied to user accounts.

Because automation accounts are rarely subject to the same identity governance or offboarding scrutiny as human users, this technique enables long-term persistence, broad data visibility, and operational concealment, facilitating continued access or covert surveillance after the subject’s departure.

A subject deliberately performs a limited technical action to determine whether a security control, detection rule, workflow, or investigative process will identify, block, or escalate the behavior. The action may appear minor in isolation, but its purpose is to validate whether a later, more serious infringement can be conducted without detection or consequence.

This behavior is distinct from the Motive ‘Boundary Testing’, which describes the subject’s motive for testing organizational tolerance. 'Testing Security Controls' is the preparatory action itself: the subject probes a specific technical or procedural control to assess whether the organization detects, prevents, or responds to the activity.

Testing may involve sending a small file to a personal email account, uploading non-sensitive material to an unapproved cloud service, installing a minor unauthorized tool, accessing a restricted repository, printing a low-value document, using an unapproved browser extension, or attempting to bypass a proxy, DLP, CASB, EDR, or identity control. The subject may then wait to see whether they are contacted by security, management, HR, or another authority, or simply confirm whether the action succeeded.

Successful testing may increase confidence, refine the subject’s method, or identify a viable path for later data exfiltration, unauthorized access, policy circumvention, sabotage, or another infringement.

Investigative Relevance

Testing security controls is often visible through low-volume, low-impact actions that precede more serious activity. The key investigative feature is not the technical action alone, but the pattern of deliberate probing, success validation, delay, and later escalation.

Investigators should assess whether the subject conducted a small-scale action before a larger attempt, repeated similar activity across different channels, or paused after the test to observe whether a response occurred. A delay between the test and later infringement may indicate that the subject was assessing organizational reaction time, alert handling, or enforcement consistency.

Example Scenarios:

• A subject emails a harmless internal document to a personal account and waits several days before attempting to transfer sensitive files.
• A subject uploads a small non-sensitive file to an unapproved cloud storage platform to test whether DLP, proxy, or CASB controls block the upload.
• A subject accesses a repository outside their normal role scope and monitors whether an access review, manager notification, or security alert follows.
• A subject installs an unauthorized browser extension to determine whether browser or endpoint controls detect unapproved extension use.
• A subject compresses, renames, encrypts, or stages a small test file before applying the same method to sensitive data.

The subject exploits circumstances for personal gain, convenience, or advantage, often without premeditation or major malicious intent. Opportunistic acts typically arise from perceived gaps in oversight, immediate personal needs, or desires, rather than long-term ideological, financial, or revenge-driven motivations.

Characteristics

• Motivated by immediate self-interest rather than deep-seated grievance or ideology.
• May rationalize actions as minor, justified, or harmless ("no one will notice," "this helps everyone," "it's not a big deal").
• Often triggered by environmental factors such as poor oversight, operational stress, or unmet personal needs.
• May escalate over time if not detected and corrected early.
• Subjects often do not view themselves as "threat actors" and may retain a positive view of their organization.
•  

Example Scenario

Senior enlisted personnel on a U.S. Navy warship collaborated to procure and install unauthorized satellite internet equipment (Starlink) to improve their onboard quality of life. Acting without command approval, they circumvented Navy IT security protocols, introducing significant operational security (OPSEC) risks. Their motive was personal convenience rather than espionage, sabotage, or financial gain.

The subject deploys destructive malware; software designed to irreversibly damage systems, erase data, or disrupt operational availability. Unlike ransomware, which encrypts files to extort payment, destructive malware is deployed with the explicit intent to delete, corrupt, or disable systems and assets without recovery. Its objective is disruption or sabotage, not necessarily for direct financial gain.

This behavior may include:

• Wiper malware (e.g. HermeticWiper, WhisperGate, ZeroCleare)
• Logic bombs or time-triggered deletion scripts
• Bootloader overwrite tools or UEFI tampering utilities
• Mass delete or format scripts (format, cipher /w, del /s /q, rm -rf)
• Data corruption utilities (e.g. file rewriters, header corruptors)
• Credential/system-wide lockout scripts (e.g. disabling accounts, resetting passwords en masse)

Insiders may deploy destructive malware as an act of retaliation (e.g. prior to departure), sabotage (e.g. to disrupt an investigation or competitor), or under coercion. Detonation may be manual or scheduled, and in some cases the malware is disguised as routine tooling to delay detection.

Destructive deployment is high-severity and often coincides with forensic tampering or precursor access based infringements (e.g. file enumeration or backup deletion).

The subject deliberately uses Virtual Private Network (VPN) technology in a manner that circumvents organizational oversight, masking the nature, destination, or content of network activity. This includes installing unapproved VPN clients, as well as reconfiguring sanctioned VPN software to route traffic through unauthorized exit nodes, personal infrastructure, or third-party services not governed by corporate policy.

By diverting traffic away from monitored pathways, the subject obstructs standard telemetry collection - evading logging of session destinations, data transfers, or identity-bound usage. This behavior frustrates forensic reconstruction, hinders real-time monitoring, and degrades the reliability of investigative artifacts. Unauthorized VPN usage is an intentional anti-forensics measure aimed at concealing potentially harmful activity behind layers of encrypted and unsanctioned transit.

A subject prepares for potential insider activity by directing an artificial intelligence (AI) agent to aggregate, organize, or transform sensitive organizational data into structured or portable formats.

This behavior occurs when an AI agent is tasked with systematically collecting information from internal repositories and consolidating it into outputs that are easier to store, review, transfer, or exploit. The agent performs bulk summarization, data normalization, or cross-repository aggregation that significantly reduces the effort required to later misuse the information.

Unlike reconnaissance activities that focus on discovering intelligence, AI Agent Data Staging focuses on operational preparation of data. The AI agent converts dispersed or complex internal information into consolidated outputs that increase its portability, usability, or accessibility outside its original context.

Examples include:

• Directing an AI agent to compile documents from multiple internal repositories into a consolidated report or briefing.
• Aggregating large volumes of internal documentation into structured summaries or datasets.
• Transforming proprietary knowledge bases or technical documentation into simplified formats suitable for external distribution.
• Generating derivative outputs that remove contextual safeguards such as system dependencies, formatting controls, or embedded metadata.
• Organizing large collections of files or records into categorized outputs intended for later retrieval or transfer.

The defining characteristic of this Sub-section is the delegated consolidation of sensitive information. The subject leverages the AI agent to perform scalable data preparation that increases the volume, portability, or usability of organizational data.

While the staged data may not yet have been transferred outside the organization, the consolidation process materially lowers the effort required to exfiltrate or exploit it. In environments where AI platforms possess broad repository visibility, this capability can significantly accelerate the preparation phase of insider activity.

A subject causes harm by directing an artificial intelligence (AI) agent to generate unauthorized internal intelligence about the organization—information the subject would not reasonably possess through legitimate role-based access or business need.

This behavior occurs when an AI agent is used to systematically query, correlate, or infer sensitive internal information by leveraging its broad integrations, indexing authority, or analytical capabilities. The resulting intelligence may reveal restricted projects, executive decisions, legal matters, acquisition activity, investigation status, architectural weaknesses, or other sensitive organizational insight.

Unlike routine search or manual browsing, AI agent internal reconnaissance enables:

• Cross-repository correlation of fragmented data.
• Inference generation from distributed signals.
• Relationship mapping between people, systems, and initiatives.
• Aggregation of intelligence from platforms the subject does not routinely access.
• Persistent monitoring for developments related to sensitive topics.

In certain enterprise deployments, authorized AI platforms possess integration-level access across knowledge bases, ticketing systems, document repositories, messaging platforms, and identity directories. A subject may deliberately leverage this aggregated visibility to extract or infer intelligence beyond their legitimate business scope.

Examples include:

• Tasking an AI agent to determine whether a confidential acquisition is underway by correlating calendar entries, procurement tickets, and legal document metadata.
• Directing an agent to summarize all references to an internal investigation across multiple repositories.
• Instructing an agent to infer likely restructuring plans based on hiring freezes, budget adjustments, and executive communications.
• Using an AI platform’s enterprise-wide indexing capability to identify sensitive project names or legal matters outside the subject’s department.

The infringement is established when the agent-generated intelligence materially exceeds legitimate business need and results in unauthorized exposure of sensitive organizational insight. The harm lies in the unauthorized acquisition of internal intelligence—particularly where that intelligence enables subsequent exploitation, trading, coercion, or strategic misuse.

The defining characteristic is not merely access, but computational synthesis. The AI agent transforms distributed internal data into actionable intelligence that the subject could not reasonably derive manually within policy boundaries.

A subject deliberately enters or remains within a restricted research facility, laboratory, prototype area, test environment, manufacturing floor, production line, formulation area, engineering workspace, or other location containing sensitive development or operational activity without authorization.

The area may expose unreleased intellectual property, prototypes, formulas, samples, research data, specialized equipment, manufacturing methods, safety-critical processes, or regulated materials. The subject’s presence may create risks to confidentiality, product integrity, safety, regulatory compliance, or operational continuity even where no additional action is observed.

A subject knowingly remains within a restricted physical area after their authorized access period has ended, or enters the area during a time when their role, assignment, escort approval, or access authorization does not permit their presence.

The subject may have legitimate daytime or task-based access to the location, but that authority does not extend to the time at which the presence occurs. Relevant behaviors include remaining after a shift or escorted visit, entering during a closed period, concealing continued presence after other personnel depart, or returning outside an approved access window.

The infringement is distinguished from accidental overstay by evidence that the subject understood the temporal restriction and deliberately remained or entered regardless.

A subject conducts a limited transfer of non-sensitive or low-value data to determine whether organizational controls detect, block, quarantine, or escalate the activity.

The subject may send a file to a personal email account, upload content to an unapproved cloud storage service, transfer data through a messaging platform, or use another external destination before attempting a larger or more sensitive transfer. They may vary the file type, size, classification, destination, compression, encryption, or transfer method to identify which conditions trigger Data Loss Prevention (DLP), web proxy, email gateway, or cloud access security controls.

The behavior may include a pause after the test while the subject waits to determine whether security personnel, management, or another authority responds. A successful test may establish a viable exfiltration route or reveal thresholds that can be avoided during a later infringement.

A subject attempts to access a system, repository, application, physical area, account, record set, or other restricted resource to determine whether the applicable access control prevents entry or generates a response.

The subject may use their own identity to request or attempt access slightly outside their normal responsibilities, test an expired or unauthorized physical identity token, navigate directly to a restricted application resource, or attempt to view information associated with another team, customer, case, or business function.

The action may be deliberately limited so that it can be described as accidental if challenged. Repeated denied attempts, access testing across several resources, or a later successful entry may indicate that the subject is mapping authorization boundaries before conducting unauthorized access or another infringement.

A subject performs a limited action on an organizational endpoint to determine whether software restrictions, endpoint security controls, browser management, device controls, or application policies prevent or report the behavior.

The subject may install a low-impact unauthorized application, add a browser extension, execute a portable utility, alter a minor endpoint setting, or attempt to run software from an unusual location. They may also connect a removable device or use a test script to establish whether application allowlisting, Endpoint Detection and Response (EDR), antivirus, mobile device management, or device-control policies are active.

The test may be designed to resemble ordinary experimentation or troubleshooting. Investigators should consider whether the subject subsequently used the same installation method, application type, device, or control weakness to support data collection, persistence, circumvention, or another infringement.

A subject performs limited network activity to determine whether firewalls, network segmentation, Domain Name System filtering, web proxies, intrusion detection systems, or protocol restrictions identify or block the connection.

The subject may attempt to reach a prohibited domain, connect to a restricted port, access another network segment, use an unapproved protocol, configure a proxy, or send a small amount of traffic through a normally restricted route. The objective is to establish whether the network path is available and whether the activity produces an alert or investigative response.

Testing may be repeated across different destinations, ports, protocols, systems, or times of day. A subject may use the results to select a less monitored network route, bypass segmentation, communicate with external infrastructure, or prepare a data-transfer channel.