Dolphin X Malware Adds AI Profiler to Rank and Prioritize High-Value Victims After Infection
Dolphin X is a new infostealer and RAT for Windows that incorporates an AI Profiler capable of scoring infected machines and ordering victims by value. The idea is to accelerate the transition from mass secret theft to subsequent compromises in accounts, cloud services, and production environments.
The novelty of Dolphin X lies not only in what it steals but in whom it decides to target first. This Windows malware, presented as both an infostealer and remote access trojan, includes a component called AI Profiler that classifies compromised machines, assigns scores, and builds a ranking so operators can concentrate time and resources on the most profitable victims.
The profiling relies on telemetry from the infected machine. This includes signals such as application usage, browser-visited domains, and installed software. From this data the attacker panel generates daily summaries and a priority order that reduces the manual triage work typically required after infection waves.
Operators first examine machines offering the greatest potential return, whether through cloud access, technical profiles, or the presence of sensitive tools. The operator panel showcases the product scope with 329 functions spread across 10 categories, a figure typical of commercial malware families that advertise extensive catalogs even before real-world campaigns are observed.
In the credential theft section, Dolphin X claims compatibility with more than 300 applications, explicitly mentioning nine Chromium and Gecko browser families, over 100 cryptocurrency wallet extensions, 65 desktop wallets, 10 password managers, and more than 30 CLI tools commonly used in cloud environments.
Among the declared targets are .env files, SSH keys, cloud access tokens, browser sessions and credentials, plus information linked to cryptocurrency wallets. The existence of the AI Profiler workflow has been verified inside the panel along with internal strings and functions related to data collection and score calculation.
However, the actual AI engine responsible for generating the valuations has not yet been confirmed. Full analysis of an active sample running on a victim machine is still needed to determine which model is used, whether external services are called, or whether the system relies on rules and weightings given an attention-grabbing name.
No active distribution campaign or specific lures have been confirmed either. This gap does not reduce the risk. The combination of secret theft and automatic prioritization can drastically shorten the time until the next stage, such as account takeover, repository access, movement into production environments, or abuse of cloud infrastructure.
Related articles
Brazilian Malware BraZetsu Uses AI to Map and Sell Compromised Corporate Networks
Researchers have identified a new Windows malware called BraZetsu developed by the Brazilian actor Exilware. The tool maps compromised corporate networks, identifies high-value systems, and prepares access for sale to other criminals. The operation primarily targets Brazil and other countries in Latin America and the Iberian Peninsula. BraZetsu is written in Python with a modular architecture and maintains WebSocket communication with its command-and-control server. It searches for banking software, ERPs, e-commerce platforms, industrial environments, development tools, backup systems, and security solutions. The malware also looks for Brazilian CNAB financial files, digital certificates, and browser history. Evidence suggests heavy use of generative AI during development, with an AI component on the server that may classify stolen files and prioritize valuable machines.
Callfuscated: Reverse Engineering a Stack-Based VM Protected by MBA, Opaque Predicates and Call-Based Jumps
A detailed technical write-up examines the HackTheBox challenge Callfuscated, which combines virtualization, mixed Boolean-arithmetic expressions, opaque predicates and junk instructions. The author first converts call-pop sequences into direct jumps by proving that register r8 is never read, then removes the resulting NOPs to obtain clean code. The resulting binary reveals a classic stack-based virtual machine whose handlers invoke heavily obfuscated functions. Triton is used to lift these functions into AST form, after which the CoBRA solver simplifies the MBA expressions into the original operations: addition, subtraction, multiplication, XOR, OR and AND. With semantics recovered, the author captures the bytecode at runtime, writes a C translator that emits readable operations, compiles the output and obtains a clean decompiled routine that checks the flag.
Backblaze Responds in Minutes While reg.ru Delays Action on Banking Trojan Infrastructure
A detailed reverse-engineering report reveals a sophisticated Android banking Trojan distributed via Telegram spam that disables antivirus apps, intercepts SMS one-time codes, and exfiltrates banking credentials. The malware uses AES-encrypted assets, a custom binary protocol over sockets, and multiple modules for screen streaming, keylogging, microphone access, and remote control. Infrastructure analysis traced the payload to an S3-compatible bucket on Backblaze, four command-and-control domains, and a VPS, with the domains registered through reg.ru. Notifications sent to Backblaze resulted in account termination within 11 minutes, while reg.ru responded only after a week with a statement refusing action without a court order. The same signing certificate appears across multiple variants, allowing rapid infrastructure rotation that keeps the campaign active despite takedown attempts. Kaspersky and CERT teams were also notified but produced no visible results within the observed period.
SLEEPWALKER Backdoor Activates on Windows via Single Custom Encrypted Packet
SLEEPWALKER is a passive Windows backdoor that stays dormant in memory until it receives one specially crafted encrypted network packet. Upon activation the implant decrypts and executes custom bytecode written in a proprietary 23-instruction language, eliminating traditional beaconing and reducing network indicators. The malware is delivered as a 64-bit DLL that impersonates dpapi.dll and is loaded via DLL side-loading into ERAAgent.exe, the executable of the ESET Management Agent used in ESET PROTECT deployments. It inspects raw traffic in promiscuous mode, supports multiple transport protocols including TCP, UDP, ICMP, SMB named pipes and VMware VMCI, and can weaken security by modifying registry values such as EveryoneIncludesAnonymous and NullSessionPipes. Configuration is protected with AES-256-CCM and the sample contains no confirmed victims or attributed infrastructure. Researchers have released YARA rules and read-only scanning utilities to detect the implant and its artifacts.