PromptRiskDBThreat intelligence atlas
AI Risk

Heterogeneous Attacks

"Heterogeneous Attacks. A closely related risk is the possibility of multiple agents combining different affordances to overcome safeguards, for which there is already preliminary evidence (Jones et al., 2024, see also Case Study 12). In this case, it is not the sheer number of agents that leads to the novel attack method, but the combination of their different abilities. This might include the agents’ lack of ind...

AI Risk7. AI System Safety, Failures, & Limitations7.6 > Multi-agent risks2 - Post-deployment

Record summary

A quick snapshot of what this page covers.

Techniques3Attack methods connected to this risk.
Mitigations5Defenses that may help with related attacks.
Domain7. AI System Safety, Failures, & LimitationsThe broad risk area this belongs to.

Risk profile

How this risk is described and categorized.

"Heterogeneous Attacks. A closely related risk is the possibility of multiple agents combining different affordances to overcome safeguards, for which there is already preliminary evidence (Jones et al., 2024, see also Case Study 12). In this case, it is not the sheer number of agents that leads to the novel attack method, but the combination of their different abilities. This might include the agents’ lack of individual safeguards, tasks that they have specialised to complete, systems or information that they may have access to (either directly or via training), or other incidental features such as their geographic location(s). The inherent difficulty of attributing responsibility for security breaches in diffuse, heterogeneous networks of agents further complicates timely defence and recovery (Skopik & Pahi, 2020)."

Domain7. AI System Safety, Failures, & Limitations
Subdomain7.6 > Multi-agent risks
Entity2 - AI
Intent1 - Intentional
Timing2 - Post-deployment
CategoryMulti-Agent Security
SubcategoryHeterogeneous Attacks

Suggested mitigations

Defenses that may help with related attacks.

AI Telemetry Logging

DeploymentMonitoring and Maintenance
LifecycleDeployment + 1 moreCategoryTechnical - Cyber

Source

Research source for this risk, when available.