AI Stack Alert High TD-2026-0016 For CISO, SOC / IR, AppSec, AI Eng AI-assisted
CERT-In warns of AI-assisted Crafty Falcon attacks on financial and crypto firms
CERT-In's finance branch (CERT-Fin) describes Crafty Falcon, a previously undocumented threat actor that uses AI and agentic AI penetration testing tools against non-banking financial companies, fintechs, payment aggregators and crypto exchanges. The actor gets in through internet-exposed UAT, staging and test environments, then moves into production using shared credentials or network paths. CERT-In says some victims were attacked again after an earlier warning because remediation was incomplete.
Why it matters
Financial and crypto firms face unauthorized transfers, wallet key theft and cloud command execution. Machine-speed, AI-generated attack scripts may slip past monitoring that is tuned to business hours and manual pacing.
Do this
- Separate UAT, staging and test environments from production. Remove shared credentials, tokens, API keys, databases and direct network paths between them.
- Alert on small-value test transactions to new beneficiaries followed by high-value transfers. Hold or review these transactions.
- Centralize cloud audit log monitoring. Alert on remote command execution through services such as AWS Systems Manager and Lambda, and on serverless code updates or new roles, users or access keys.
- Monitor LLM API usage and keys. Investigate LLM API traffic from servers or non-development hosts.
- Keep 24/7 monitoring and lower alert thresholds for weekends, holidays and other low-monitoring windows. Patch internet-facing PHP and Node.js applications quickly, prioritizing file-upload and deserialization flaws, and use WAF virtual patching as an interim measure.
Signals
- NoCISA KEV
- 36/100Priority
- 80%Confidence
Evidence
“Crafty Falcon is a previously undocumented threat actor that uses a range of AI and agentic AI penetration testing tools to breach the systems of its victims.”
“Typically, Crafty Falcon gains initial access through internet-exposed non-production environments, such as user acceptance testing (UAT), staging and testing environments.”
“Despite undertaking remediation after receiving this previous warning, these financial organizations were attacked again”
“LLM API traffic originating from servers or non-development hosts may indicate an attack infrastructure or planted operator console”
Sources
Affected / related
Crafty Falcon, CERT-In, ARTEX, PentAGI, CyberStrikeAI, Cairn, AWS Systems Manager, AWS Lambda
Published 2026-11-05 14:25 UTC · updated 2026-11-10 00:24 UTC · AI-assisted summary, checked against quoted sources · How we score