Fracture — DECIDE Loop Corruption

Corrupt how the platform thinks.

Classification

Attribute	Value
Type	Active — alters target state
Target	DECIDE loop
Effect	Platform makes decisions that serve the defender
Phases	Primarily Phase E, informed by Phase F (pre-engagement RE)
Hardening Viability	Tier 0-2: viable where decision logic is understood. Tier 3: requires novel research

Definition

Fracture corrupts the platform’s decision-making. Where Mirage feeds false data to SENSE, Fracture targets the logic that interprets that data. The platform may perceive the world accurately but reach incorrect conclusions — or its decision thresholds may be manipulated to trigger state transitions at attacker-chosen times.

Fracture operates at two scales:

Single-platform decision manipulation — exploiting the autonomous decision logic of an individual platform
Multi-platform swarm logic corruption — poisoning coordination protocols that enable collective behavior

Single-Platform Techniques

Applicable to any platform with documented or reverse-engineered decision logic. Validated through failsafe mapping (Phase F pre-engagement / teardown Phase 9).

Failsafe Threshold Manipulation

Every autonomous platform has conditional logic: IF condition X, THEN action Y. Fracture manipulates the inputs evaluated against those thresholds.

Threshold	Manipulation	Effect
Altitude floor (e.g., 40m)	GPS altitude spoof below threshold	Forces VTOL motor engagement
VTOL-assisted altitude (e.g., 30m)	GPS altitude spoof below 30m	Forces full VTOL stabilization — max power draw
Roll limit (e.g., >30°)	GPS spoof induces navigation corrections exceeding attitude limit	Forces VTOL stabilization failsafe
Stall speed (e.g., <18 m/s)	GPS velocity spoof corrupts airspeed failover	Forces VTOL motor engagement
Battery voltage (e.g., <41V)	Indirect — force high-consumption state via other thresholds	Forces Q-LAND
Airspeed-ground speed delta (e.g., >20 m/s)	GPS ground speed spoof	Forces airspeed failover

The distinction between Fracture and Mirage: the GPS spoof is the delivery mechanism (Mirage). The failsafe threshold crossing is the effect (Fracture). In practice, Mirage and Fracture are often chained.

State Machine Exploitation

Force the platform through unintended state transitions:

AUTO → RTL by triggering the appropriate failsafe
Fixed-wing cruise → VTOL hover by triggering attitude/altitude thresholds
Armed → disarmed by exploiting pre-arm check logic in reverse

Each forced transition consumes time, battery, and operator attention.

Geofence Exploitation

Spoof the platform’s position to appear outside its geofence boundary. The platform’s own safety feature triggers geofence-breach behavior (typically RTL or landing).

Multi-Platform Techniques (Forward-Positioned)

These target platforms coordinating via mesh networking, swarm protocols, or shared situational awareness. Emerging capability area.

Swarm Logic Poisoning

Inject false peer data into swarm coordination protocols — false position reports, false task assignments, false obstacle detections. Corrupting collective input corrupts collective behavior.

Autonomous Negotiation Corruption

Disrupt task allocation by injecting conflicting assignments or spoofing task completion messages.

Coordination Clock Disruption

Introduce timing offsets between platforms. If platform A is at T+0 and platform B at T+200ms, coordinated behavior degrades.

Fracture vs. Hardening

Hardening	Impact on Fracture
Authenticated sensor inputs	Reduces delivery vectors
Redundant decision inputs	Single-input manipulation may be detected
Adaptive thresholds	Fixed threshold exploitation becomes unreliable
Authenticated swarm protocols	Swarm poisoning requires defeating authentication
Byzantine fault tolerance	Requires corrupting N+1 peers

Fracture remains viable at higher hardening tiers than Override because decision logic complexity creates a larger attack surface than communication protocol security. Even authenticated, encrypted systems must still make decisions — and those decisions have exploitable logic.