Back to the MIT repository
2. Privacy & Security3 - Other

Risks from network interconnectivity

The interconnectedness of AI networks can create vulnerabilities, where issues in one part of the network can have cascading effects across the system.

Source: MIT AI Risk Repositorymit1085

ENTITY

3 - Other

INTENT

3 - Other

TIMING

3 - Other

Risk ID

mit1085

Domain lineage

2. Privacy & Security

186 mapped risks

2.2 > AI system security vulnerabilities and attacks

Mitigation strategy

1. Implement an AI-native governance model utilizing "circuit breakers." Design controls to automatically slow, suspend, or redirect system behavior upon the breach of predefined thresholds (e.g., error rates or security signals) to prevent the uncontrolled propagation of an initial failure across interconnected AI networks. 2. Model and simulate network interdependencies for reinforcement. Employ advanced AI techniques, such as Graph Neural Networks (GNNs) and optimization algorithms, to map the complex interdependencies within the AI network, simulate failure scenarios, and identify critical nodes for the optimal placement of redundant resources to enhance systemic resilience. 3. Establish cross-organizational risk-based controls and shared protocols. Extend risk management beyond internal defenses to address supply chain dependencies and correlated failure modes by mandating shared alerting mechanisms, coordinated mitigation plans, and clear escalation protocols across all integrated systems, suppliers, and users.