Cybersecurity ·

India Warns of AI-Assisted “Crafty Falcon” Attacks on Crypto and Fintech Firms

India’s financial cybersecurity response team says an AI-assisted threat actor is targeting crypto exchanges and fintech firms, abusing payment APIs to steal funds.

India Warns of AI-Assisted “Crafty Falcon” Attacks on Crypto and Fintech Firms

A new name for a financial cyber campaign

India’s Computer Security Incident Response Team has warned financial companies about a threat actor it calls “Crafty Falcon,” describing an AI-assisted campaign targeting non-bank lenders, fintech firms, payment providers and cryptocurrency exchanges. According to the advisory, the group uses machine intelligence to survey targets and can operate at high speed, including by abusing the victim’s own payment and disbursement application programming interfaces.

The description suggests the threat is not limited to stealing login details. If attackers can manipulate trusted payment functions, they may cause systems to send money through processes that appear routine. Crypto businesses may be especially exposed because transfers can be fast and difficult to reverse once wallet keys or transaction approvals are compromised.

What “AI-assisted” does and does not mean

The label indicates that investigators believe artificial intelligence is part of the actor’s workflow. It does not automatically mean the group has developed autonomous hacking software or that AI alone caused the breaches. AI can assist with reconnaissance, message personalization or analysis, while the attack still depends on access, software weaknesses and operational mistakes.

The advisory describes a serious escalation but does not establish the group’s nationality or state sponsorship. Attribution requires evidence beyond a name and technical pattern. A group’s tools can be copied, rented or deliberately made to resemble another actor’s methods.

Defenses and unanswered questions

CERT-In advises affected organizations to examine API security and avoid relying only on IP allow-lists for shared third-party services. Stronger measures can include mutual TLS, signed requests, narrowly scoped API keys, robust logging and monitoring for unusual payment behavior. Companies should also rehearse how they would disable a compromised integration without shutting down critical services for customers.

Crypto users can reduce exposure by enabling strong authentication and separating long-term holdings from wallets connected to everyday applications. But individual precautions cannot substitute for institutional controls on payment systems.

How quickly can regulators and firms adapt when attackers automate parts of the process? And can organizations protect transaction APIs without making digital payments less reliable? The advisory makes those questions immediate, while the campaign’s full scale and attribution remain under investigation.