WORLDTECH NEWS Global technology intelligence.Contact
Cybersecurity SINGLE SOURCE

Bazuka Android Trojan Combines Keylogging With SMS Code Interception

Analyst seen from behind facing a wall of blurred monitoring screensAI illustration
WORLDTECH illustration · AI-generated (Canva)Google logo shown for identification only; no affiliation with or endorsement of WORLDTECH is implied.

A newly disclosed Android banking Trojan called Bazuka captures banking logins and intercepts SMS verification codes, and one outlet reports that its operators rely on large language models to decide which infected handsets to pursue.

What the framework does

Bazuka is a modular Android banking Trojan that captures banking credentials. According to the available report, it also intercepts text messages carrying verification codes, data that can help an attacker get past a login check.

The most distinctive element, according to the same account, is the use of large language models by its operators to help rank compromised devices by how closely they should be followed.

How the framework came to light

ThreatFabric identified the framework in late October 2026 during an investigation into an incident involving an NFC card relay application.

The report adds that Bazuka combines a dropper with further components, although the verified material does not describe how that combination is built.

What remains unconfirmed

All details come from a single outlet, GBHackers, which relays the findings. The available material does not state how many devices have been affected, which banks or apps are targeted, or how the large language models are used in practice.

Open questions

  • How many Android devices have been infected by Bazuka?
  • Which banking applications does the framework target?
  • How do the operators use large language models in practice?

Key facts

  • Bazuka is a modular Android banking Trojan that captures banking credentials.
  • It intercepts SMS messages that carry verification codes.
  • Its operators reportedly use large language models to help prioritise compromised devices.
  • ThreatFabric identified the framework in late October 2026.
  • The discovery arose from an investigation into an incident involving an NFC card relay application.

Sources & evidence