On 4 November 2026, India's national cybersecurity agency published advisory CIAD-2026-0058 with a severity rating of Critical. It named an AI-assisted threat actor — CRAFTY FALCON — that moves money by calling victims' own payment and disbursement APIs.
This article separates what is officially confirmed from what has only been reported, because the story moved fast across several publications and at least one central figure was formally denied by the bank named in it.
What CERT-In's advisory actually says
- CRAFTY FALCON is described as "an AI-assisted financial threat actor that surveys its targets with machine intelligence, operates at machine speed and strikes by invoking the victim's own payment and disbursement APIs."
- Primary targets: NBFCs, digital lending platforms, fintechs, payment aggregators, PSPs, crypto exchanges.
- Named tooling: ARTEX, PentAGI, Cairn, CyberStrikeAI — all publicly available.
- Status: active and ongoing. Entities that had already remediated were deliberately re-targeted, including after incomplete remediation.
- Attacks were timed for holidays, weekends and other low-monitoring windows.
A parallel advisory from the Indian Cybercrime Coordination Centre (I4C) described the same pattern: attacks "carried out with the help of Artificial Intelligence (AI) agents leading to machine speed compromise," and recommended daily and per-transaction limits, beneficiary-name verification, and watching for unusual transaction speeds.
Which incidents are actually confirmed
- Two Mumbai-based NBFCs lost roughly $2 million across 700+ unauthorised instant transfers from ICICI Bank accounts on 8–9 October 2026 (police complaint, per Reuters).
- State Bank of India: 3,897 customer records with names, phone numbers and addresses found on the dark web.
- Spandana Sphoorty Financial reported unauthorised debits; AU Small Finance Bank is probing a potential data leak.
- Yes Bank's own confirmed incident: a ₹2.54 crore forex-card breach affecting ~5,000 customers, with 688 unauthorised attempts blocked, for which RBI summoned bank officials.
A 13 November report alleged a $3.4 million loss at Yes Bank. Yes Bank denied it the same day, calling the claims factually incorrect. CERT-In also stated it did not say which incidents were linked to CRAFTY FALCON. Both facts are noted in the article rather than repeated as established.
The open-source tooling behind the campaign
Elastic Security Labs read the source of five agentic offensive frameworks (ARTEX, CyberStrikeAI, Cairn, PentAGI, HexStrike). Findings:
- Over 60,000 GitHub stars between them; confirmed use in real intrusions, not only red-team engagements.
- None enforces authorization in code. Three assert it only in the prompt, and the only default target restrictions are blocklists covering government and education domains.
- ARTEX operates in bursts of roughly 2–25 seconds; a full operation can span hours, so detection needs a long lookback window.
- The frameworks change route when blocked, but their actions do not — behaviour-based detection outlasts indicator-based blocking.
Regulatory response
On 30 October 2026 the Reserve Bank of India increased scrutiny of banks' cybersecurity preparedness, seeking proof of compliance with existing rules and details on AI-risk preparedness. I4C's 4 November advisory added the operational safeguards above.
Defensive checklist for Indian finance teams
- Inventory every API that can move money, including partner and forgotten endpoints.
- Replace static tokens in source code and CI configuration with short-lived, scoped credentials.
- Set daily and per-transaction limits; verify beneficiary names programmatically.
- Alert on unusual transaction velocity; confirm thresholds fire on bursts measured in seconds.
- Staff monitoring for weekends and holidays — that is when these attacks were scheduled.
- Anchor detections on behaviour, not indicators alone.
- Re-test the full attack chain after any fix; partial remediation is what got entities re-targeted.
- Know the reporting paths: incident@cert-in.org.in for technical logs, the CFCFRMS portal for bulk fraud.
What this article is not
It is not an incident-response report or a security assessment of any specific institution, and it does not claim that any named bank suffered an AI-driven breach. We have not independently verified any loss figure against a regulatory filing, and we have not reproduced any of the named attack tooling.
Read the full post on PyBankers: https://pybankers.com/blog/ai-financial-fraud-india-crafty-falcon-advisory-2026
