Agentic Design

Patterns

Multimodal Attacks

Cross-modal exploitation and modality-specific attack techniques

10
Techniques
3
medium
medium Complexity
7
high
high Complexity

Available Techniques

🖼️

Image-Based Prompt Injection

(IBPI)
medium

Embedding malicious text instructions or prompts within images to bypass text-based content filters and inject harmful directives through the visual modality.

Key Features

  • •Hidden text in images
  • •Visual prompt injection
  • •OCR exploitation

Primary Defenses

  • •Image content analysis
  • •OCR output sanitization
  • •Cross-modal validation

Key Risks

Content filter bypassHidden malicious instructionsCross-modal security evasionStealth attack vectors
🔀

Cross-Modal Confusion Attack

(CMCA)
high

Exploiting inconsistencies or conflicts between different input modalities to confuse the AI system and bypass security controls or trigger unintended behaviors.

Key Features

  • •Modality conflict exploitation
  • •Contradictory input injection
  • •Priority manipulation

Primary Defenses

  • •Cross-modal consistency validation
  • •Modality agreement requirements
  • •Conflict detection and rejection

Key Risks

Security control bypass through confusionInconsistent behavior exploitationFusion mechanism vulnerabilitiesPriority manipulation
🎵

Audio Adversarial Examples

(AAE)
high

Crafting audio inputs with imperceptible perturbations that cause speech recognition or audio processing systems to misinterpret commands or bypass security measures.

Key Features

  • •Imperceptible audio perturbations
  • •Speech recognition manipulation
  • •Command misinterpretation

Primary Defenses

  • •Audio perturbation detection
  • •Speech pattern validation
  • •Multi-model audio verification

Key Risks

Unauthorized voice commandsSpeech recognition bypassStealth audio attacksVoice-activated system compromise
🎬

Video Manipulation & Injection

(VMI)
high

Manipulation of video streams or recorded content to inject malicious visual sequences, subliminal frames, or adversarial patterns that compromise video understanding systems.

Key Features

  • •Frame injection
  • •Subliminal content insertion
  • •Temporal attack patterns

Primary Defenses

  • •Frame-by-frame validation
  • •Temporal consistency checks
  • •Subliminal content detection

Key Risks

Subliminal instruction injectionVideo content manipulationTemporal security bypassMotion-based attacks
📡

Sensor Data Poisoning

(SDP)
high

Manipulation of sensor inputs (IoT devices, environmental sensors, biometric readers) to feed false data to AI systems and compromise decision-making in autonomous systems.

Key Features

  • •Sensor input manipulation
  • •Environmental data falsification
  • •Biometric spoofing

Primary Defenses

  • •Sensor data validation
  • •Multi-sensor verification
  • •Anomaly detection algorithms

Key Risks

Autonomous system compromiseSafety-critical failuresEnvironmental misinterpretationBiometric system bypass
🔓

Modality-Specific Jailbreaking

(MSJ)
medium

Bypassing content filters and safety measures by exploiting weaknesses in specific modality processing, using less-protected input channels to circumvent text-based safeguards.

Key Features

  • •Modality-specific filter bypass
  • •Weak channel exploitation
  • •Alternative input abuse

Primary Defenses

  • •Unified safety filters across modalities
  • •Equivalent protection per channel
  • •Cross-modal content analysis

Key Risks

Safety measure bypassContent filter circumventionHarmful content generationInconsistent protection
🎯

Embedding Space Manipulation

(ESM)
high

Crafting inputs across multiple modalities that occupy similar positions in embedding space to confuse similarity matching, retrieval, or classification systems.

Key Features

  • •Embedding collision creation
  • •Similarity exploitation
  • •Retrieval manipulation

Primary Defenses

  • •Embedding space validation
  • •Multi-modal consistency checking
  • •Semantic verification

Key Risks

Retrieval system manipulationClassification errorsSimilarity matching failuresContent mismatch exploitation
↔️

Cross-Modal Transfer Attack

(CMTA)
high

Crafting adversarial examples in one modality that successfully transfer to compromise other modalities, exploiting shared representations in multimodal models.

Key Features

  • •Transferability exploitation
  • •Shared representation attacks
  • •Cross-modal perturbations

Primary Defenses

  • •Modality-specific processing
  • •Transfer detection mechanisms
  • •Independent validation per modality

Key Risks

Multi-modality compromiseUniversal attack patternsCascading failuresShared vulnerability exploitation
🚪

Multimodal Backdoor Attack

(MBA)
high

Inserting backdoors that activate only when specific combinations of inputs across multiple modalities are present, creating stealthy trigger-based compromises.

Key Features

  • •Multi-modal trigger conditions
  • •Combination-based activation
  • •Stealthy backdoor insertion

Primary Defenses

  • •Training data validation
  • •Backdoor detection algorithms
  • •Multi-modal integrity checks

Key Risks

Persistent hidden compromisesDifficult detectionTrigger-based exploitationStealthy attacks
⚖️

Modality Prioritization Exploitation

(MPE)
medium

Exploiting the system's prioritization or weighting of different input modalities to bypass security controls by manipulating lower-priority channels.

Key Features

  • •Priority order exploitation
  • •Weight manipulation
  • •Low-priority channel abuse

Primary Defenses

  • •Balanced modality processing
  • •Equal validation across channels
  • •Dynamic priority adjustment

Key Risks

Low-priority channel exploitationUnbalanced security coverageAttention mechanism abusePriority-based bypass

Ethical Guidelines for Multimodal Attacks

When working with multimodal attacks techniques, always follow these ethical guidelines:

  • • Only test on systems you own or have explicit written permission to test
  • • Focus on building better defenses, not conducting attacks
  • • Follow responsible disclosure practices for any vulnerabilities found
  • • Document and report findings to improve security for everyone
  • • Consider the potential impact on users and society
  • • Ensure compliance with all applicable laws and regulations

AI Red Teaming

closed

Loading...