Exposed CyberStrikeAI session logged 221 unapproved tool runs
First merged summary.
Huntback found an exposed CyberStrikeAI workspace where 221 tool calls ran after approval enforcement was switched off.
Huntback Research examined a 6.69 MB SQLite snapshot from an exposed CyberStrikeAI workspace used in a November 16, 2026 offensive testing session. Saved configuration showed approval enforcement off by 09:48:15 UTC. Of 240 runtime tool calls in one conversation, 221 had persistent execution IDs, all from the main orchestrator; 209 completed and 12 failed. A person still directed the open-source agent in chat, but four approval rejections came from an automated reviewer, not a human. Huntback did not identify the operator and says the records do not prove a shell, internal access, stolen funds, a data-lake breach, or destructive database changes.
- One conversation held 16 messages and 240 runtime tool calls.
- Approval enforcement was off at 09:48:15 UTC on November 16, 2026.
- 221 calls had execution IDs; 209 completed and 12 failed.
- The locally recorded model identifier was deepseek-v4-pro.
- Artifacts were retrieved from 38.76.169[.]207 on TCP port 8010.
- Huntback found no proof that 10,000 records or 260 unique nodes were taken.
Timeline
dated events in the reporting · oldest first- Nov 16, 2026CyberStrikeAI testing session ran with approval enforcement off. · GBHackers
What each outlet adds
beyond the shared facts- GBHackersSnapshot size, message count, artifact host and port, and Ocean Protocol caveat.
- Cyber Security NewsModel identifier, unidentified operator, and no proof of large data theft.
- Products
- CyberStrikeAI
- Organizations
- huntbackHuntback ResearchOWASP
- AI models
- deepseek-v4-pro
Coverage
oldest first · one row per article · UTCCyberStrikeAI ran 221 tools after approval enforcement was turned off
Huntback found an exposed session database, not a confirmed intrusion, shell, or data-lake breach.
Exposed CyberStrikeAI session ran 221 tool calls with approval gate off
Huntback found the approval gate disabled and could not show the agent bypassed it or confirm testing was authorized.