Record summary
A quick snapshot of what this page covers.
Techniques3Attack methods connected to this risk.
Mitigations0Defenses that may help with related attacks.
Domain4. Malicious Actors & MisuseThe broad risk area this belongs to.
Risk profile
How this risk is described and categorized.
"Cyber risks, especially in the context of cyber offense, are an existing threat that may be exacerbated by AI. [108] demonstrated that teams of LLM agents can exploit zero-day vulnerabilities when given a description of the vulnerability and toy capture-the-flag problems. While cyber risks are not typically regarded as catastrophic, [3] argues that cyberwarfare is an underappreciated risk that poses a credible threat of catastrophic harm."
Domain4. Malicious Actors & Misuse
Subdomain4.2 > Cyberattacks, weapon development or use, and mass harm
Entity3 - Other
Intent1 - Intentional
Timing2 - Post-deployment
CategoryCyber offense
Subcategoryn/a
Suggested mitigations
Defenses that may help with related attacks.
No propagated mitigations. No defense is available through the connected attack methods.
Source
Research source for this risk, when available.
Included resource
Dimensional Characterization and Pathway Modeling for Catastrophic AI Risks
Original source
MIT AI Risk Repository
Open the public repository used for AI risk records and taxonomy fields.