Tenable One Adversary View, powered by Claude Mythos, reaches general availability for attack-path prioritization
Tenable announced general availability of Tenable One Adversary View, a feature of Tenable Hexa AI that uses Anthropic's Claude Mythos model to reason over existing scan data. It chains vulnerabilities into attack paths and ranks fixes by the number of paths each one closes. This is a product announcement, not a threat or an incident, and no new vulnerability, exploitation or threat actor is reported. The article mentions EternalBlue (the SMB exploit behind WannaCry) only as an example finding in a demo. No external research was done for this record, so it relies on the article alone.
Detalhes técnicos
- It is a point-in-time, on-demand analysis. It complements the rule-based continuous attack path analysis.
- It builds an asset graph from existing scan evidence (for example NETSTAT_CONN output and CVEs), with no new sensors or agents.
- Output includes ranked chains, choke points, patch and configuration fixes, Jira tickets, dashboards and PDF reports.
- Prerequisites: a Tenable One Foundation or Advanced license, "Frontier AI Model Access" enabled in Hexa AI settings, a role with access, and scans run with "Override normal verbosity" and "Report as much information as possible".
- In the demo, one firewall rule blocking SMB from the internet closed 32 SMB connections.
Linha do tempo
-
Tenable One Adversary View, powered by Claude Mythos, reaches general availability for attack-path prioritization
ler a análise desta fonte ocultar
Tenable announced general availability of Tenable One Adversary View, a feature of Tenable Hexa AI that uses Anthropic's Claude Mythos model to reason over existing scan data. It chains vulnerabilities into attack paths and ranks fixes by the number of paths each one closes. This is a product announcement, not a threat or an incident, and no new vulnerability, exploitation or threat actor is reported. The article mentions EternalBlue (the SMB exploit behind WannaCry) only as an example finding in a demo. No external research was done for this record, so it relies on the article alone.
Technical Details
- It is a point-in-time, on-demand analysis. It complements the rule-based continuous attack path analysis.
- It builds an asset graph from existing scan evidence (for example NETSTAT_CONN output and CVEs), with no new sensors or agents.
- Output includes ranked chains, choke points, patch and configuration fixes, Jira tickets, dashboards and PDF reports.
- Prerequisites: a Tenable One Foundation or Advanced license, "Frontier AI Model Access" enabled in Hexa AI settings, a role with access, and scans run with "Override normal verbosity" and "Report as much information as possible".
- In the demo, one firewall rule blocking SMB from the internet closed 32 SMB connections.