Kai adds External Attack Surface Management to Autonomous Defense Platform and launches Kai Labs
Kai announced External Attack Surface Management (EASM) as a new capability on the Kai Autonomous Defense Platform and the launch of Kai Labs, an expanded research organization. EASM unifies internal and internet-facing exposures into a single dataset and remediation plan, aiming to let teams see exposures the way attackers do and close them through one workflow. Key capabilities mentioned include discovery and attribution of internet-facing assets, detection of exposed AI infrastructure (LLM endpoints, MCP servers, vector databases), API endpoint visibility, and attacker-style investigation. Kai Labs, led by Kylie McClanahan, PhD, will focus on research areas where AI and security intersect, including offensive learning, autonomous action, decision models, trusted autonomy, and security context.
Detalhes técnicos
- EASM discovery and attribution: discovers internet-facing assets tied to an organization’s domains and aims to surface assets missing from CMDBs and internal inventories (helpnetsecurity.com).
- Exposed AI infrastructure detection: specifically calls out LLM endpoints, MCP servers, and vector databases as targets of discovery.
- API visibility: parses published API definitions into individual endpoints for inspection.
- Workflow integration: external exposures are joined to the same asset records and queues as internal findings so remediation follows a single ownership and workflow model.
- Kai Labs research focus: offensive learning, autonomous action, decision models, trusted autonomy, and security context; outputs will be published via blogs, talks, and papers (helpnetsecurity.com).
Linha do tempo
-
Kai adds External Attack Surface Management to Autonomous Defense Platform and launches Kai Labs
ler a análise desta fonte ocultar
Kai announced External Attack Surface Management (EASM) as a new capability on the Kai Autonomous Defense Platform and the launch of Kai Labs, an expanded research organization. EASM unifies internal and internet-facing exposures into a single dataset and remediation plan, aiming to let teams see exposures the way attackers do and close them through one workflow. Key capabilities mentioned include discovery and attribution of internet-facing assets, detection of exposed AI infrastructure (LLM endpoints, MCP servers, vector databases), API endpoint visibility, and attacker-style investigation. Kai Labs, led by Kylie McClanahan, PhD, will focus on research areas where AI and security intersect, including offensive learning, autonomous action, decision models, trusted autonomy, and security context.
Technical Details
- EASM discovery and attribution: discovers internet-facing assets tied to an organization’s domains and aims to surface assets missing from CMDBs and internal inventories (helpnetsecurity.com).
- Exposed AI infrastructure detection: specifically calls out LLM endpoints, MCP servers, and vector databases as targets of discovery.
- API visibility: parses published API definitions into individual endpoints for inspection.
- Workflow integration: external exposures are joined to the same asset records and queues as internal findings so remediation follows a single ownership and workflow model.
- Kai Labs research focus: offensive learning, autonomous action, decision models, trusted autonomy, and security context; outputs will be published via blogs, talks, and papers (helpnetsecurity.com).