India’s Computer Emergency Response Team (CERT-In) issued a critical warning on November 4, 2026, about an active campaign targeting banks and other financial-sector organizations. The advisory names the threat actor “CRAFTY FALCON” and attributes AI-assisted offensive tooling to the campaign. The technical claims are CERT-In’s assessment; no public independent forensic reconstruction has been released.
What is CRAFTY FALCON?
CERT-In Advisory CIAD-2026-0058, dated November 4, updates an earlier October 9 warning about attacks exploiting payment and disbursement APIs. CERT-In’s Computer Security Incident Response Team–Finance (CSIRT-Fin) says incidents have continued, including renewed attacks on some entities after incomplete remediation.
The affected categories identified in the advisory include non-banking financial companies (NBFCs), digital lenders, fintech businesses, payment aggregators, payment service providers, cryptocurrency exchanges and virtual digital asset service providers. CERT-In does not publish a verified victim count or financial-loss total.
How are financial institutions being attacked?
According to CERT-In, the reported incidents involve several recurring paths:
- Cloud management abuse: attackers use cloud management or serverless services to execute commands and deploy malicious files.
- Staging-to-production movement: exposed testing environments become entry points when credentials, tokens or network access are shared with production.
- Payment API manipulation: altered client packages and unauthorized requests bypass controls that assume requests originate from legitimate applications.
- Small test transfers: attackers reportedly test payment routes with micro-transactions, including ₹1, before attempting larger transfers.
- Cryptocurrency infrastructure: the advisory reports wallet-key theft and unauthorized fund transfers at some targeted services.
These are patterns described in the government advisory, not evidence that every listed technique occurred at every affected organization.
What role does AI allegedly play?
CERT-In assesses that the attackers use publicly available AI-enabled offensive-security tools, naming ARTEX, PentAGI, Cairn and CyberStrikeAI. It describes rapid creation and adaptation of attack scripts and coordinated exploitation attempts.
Important distinction: The advisory is primary evidence that CERT-In made these assessments; it is not independently verified proof that a particular AI system autonomously conducted a particular theft. The publicly available advisory does not disclose the full indicators of compromise or detailed forensic evidence needed to independently establish the AI tools’ operational roles.
What should banks, fintechs and payment providers do?
CERT-In recommends heightened monitoring and stronger controls across the systems attackers reportedly targeted. The most actionable measures include:
- Maintain 24/7 monitoring, including weekends and holidays, with alerts for unusual cloud administrative actions.
- Separate staging and testing environments from production, including credentials, tokens, databases and network paths.
- Validate transactions and business rules on the server, rather than trusting the mobile or web client; consider signed requests and replay protections.
- Detect unusual payment patterns, including small transfers to new beneficiaries followed by high-value transactions.
- Replace reliance on shared third-party IP allowlists with identity-based authentication and tightly scoped credentials.
- Rotate suspected exposed secrets, investigate persistence and verify complete remediation after an incident.
- Protect cryptocurrency wallet keys using hardware-backed or offline custody where appropriate, and require strong withdrawal approvals.
These steps reflect the advisory’s recommendations; implementation should be tailored to each institution’s architecture and regulatory obligations.
What remains unknown?
CERT-In has not publicly identified victims, released aggregate losses, named the campaign operators or supplied unrestricted forensic indicators. Its restricted technical guidance has been shared with relevant entities. The precise autonomy and effectiveness of the AI tools also remain unverified in the public record.
Why the warning matters
Financial-sector attacks can progress from a weak cloud or testing environment to transaction infrastructure if credentials and trust boundaries are poorly separated. CERT-In’s escalation signals an active threat to institutions handling payments and digital assets, while also highlighting why AI-related attribution should be reported carefully rather than treated as proven by a headline alone.
Official sources
- CERT-In Advisory CIAD-2026-0058: Escalation of AI-Driven Attacks on BFSI Entities (November 4, 2026)
- CERT-In 2026 advisories index, including the October 9 predecessor advisory CIAD-2026-0046
Editorial note: This report is based on CERT-In’s public advisory. Victim totals, financial losses and the detailed AI attack chain have not been independently established.
