Virtual SOC ยท Live
Command Deck
10
Alerts
7
Crit + High
1
Investigations
1
Ready to work
AI Securitynovel
10 signals- 75Support bot returning other customers' dataRAG support assistant leaked another tenant's records after a crafted message.HIGHFINDINGReady
- 73Fraud model accuracy silently degradingFraud-detection model started passing a class of fraudulent transactions after retraining.HIGHFINDINGPreview
- 94Autonomous agent making unexpected outbound callsAn internal LLM agent host is beaconing to an unknown external endpoint.CRITFINDINGPreview
- 73Model reproducing internal source code verbatimThe internal code-assistant emits proprietary code it should never surface.HIGHFINDINGPreview
- 54LLM provider bill spiking overnightA leaked LLM API key is being used for someone else's workload.MEDINVSTGPreview
- 87Embeddings service exposed without authThe vector database is internet-reachable and returns PII-laden chunks.CRITFINDINGPreview
- 75Voice-cloned CFO authorizing an urgent wireFinance received a convincing voice call and email chain impersonating the CFO.HIGHFINDINGPreview
- 48Sensitive documents uploaded to an unsanctioned AI toolCASB flags employees pasting internal docs into an unapproved AI writing SaaS.MEDFINDINGPreview
- 46Customer chatbot coaxed past its guardrailsUsers are sharing a jailbreak that makes the public chatbot ignore policy.MEDFINDINGPreview
- 92Automation agent granting itself adminAn ops agent used a legit tool to add itself to an admin group.CRITFINDINGPreview
How's the SOC so far?