# Fake AI Tool Attacks on SMBs: 33,300 Cyberattacks Masquerading as ChatGPT, Copilot, Claude in Early 2026

> Kaspersky detected over 33,300 cyberattacks targeting small-to-medium businesses in the first four months of 2026, with attackers masquerading as popular AI tools (ChatGPT, Microsoft Copilot, Claude, Gemini). This represents a five-fold increase from 2025 attack volumes. Over 1,100 unique Trojware samples were identified, capable of downloading and executing additional malware, with confirmed data exfiltration and credential theft functionality.

- **Published:** 2026-06-28T00:00:00Z
- **Last reviewed:** 2026-06-28T00:00:00Z
- **Canonical:** https://intel.threadlinqs.com/threat/TL-2026-0993
- **ID:** TL-2026-0993
- **Severity:** CRITICAL
- **Category:** MALWARE
- **Status:** ACTIVE
- **Detections:** 9 · **IOCs:** 20 (full data via the Threadlinqs MCP server — Purple tier)

## Description

A coordinated wave of trojware campaigns targeting SMBs worldwide has exploited the explosive adoption of AI tools by creating convincing fake installers and promotional materials for ChatGPT, Microsoft Copilot, Claude, Google Gemini, and other AI platforms. The attack chain begins with phishing emails containing malicious links to fake download pages or direct trojan attachments disguised as AI tool installers. Victims are tricked into executing trojware binaries, which establish persistence through registry modifications and scheduled tasks. Once deployed, the trojware samples download and execute secondary malware including information stealers, credential dumpers, backdoors (AsyncRAT, RemotePC, Remcos variants), and potentially ransomware payloads. Analysis of the 1,100+ unique samples revealed polymorphic packing, obfuscation, and anti-analysis techniques designed to evade antivirus detection. The trojware families primarily target Windows SMB endpoints to harvest credentials, steal sensitive files (financial documents, intellectual property, customer data), monitor user activity via keylogging and screen capture, and establish persistent remote access. C2 infrastructure analysis shows fallback mechanisms with hardcoded domains rotating through compromised hosting providers and free DNS services. The campaign demonstrates sophisticated social engineering, leveraging AI tool popularity at a time when many SMBs are experimenting with these platforms. Kaspersky telemetry indicates active targeting of industries including healthcare, financial services, legal, accounting, manufacturing, and retail sectors.

## MITRE ATT&CK

- T1566 Phishing
- T1190 Exploit Public-Facing Application
- T1189 Drive-by Compromise
- T1204 User Execution
- T1059 Command and Scripting Interpreter
- T1106 Native API
- T1559 Inter-Process Communication
- T1547 Boot or Logon Autostart Execution
- T1053 Scheduled Task/Job
- T1543 Create or Modify System Process
- T1548 Abuse Elevation Control Mechanism
- T1134 Access Token Manipulation
- T1036 Masquerading
- T1027 Obfuscated Files or Information
- T1140 Deobfuscate/Decode Files or Information
- T1564 Hide Artifacts
- T1218 System Binary Proxy Execution
- T1070 Indicator Removal
- T1685 Disable or Modify Tools
- T1056 Input Capture
- T1555 Credentials from Password Stores
- T1003 OS Credential Dumping
- T1110 Brute Force
- T1087 Account Discovery
- T1217 Browser Information Discovery
- T1083 File and Directory Discovery
- T1082 System Information Discovery
- T1570 Lateral Tool Transfer
- T1115 Clipboard Data
- T1113 Screen Capture
- T1125 Video Capture
- T1123 Audio Capture
- T1114 Email Collection
- T1560 Archive Collected Data
- T1041 Exfiltration Over C2 Channel
- T1048 Exfiltration Over Alternative Protocol
- T1071 Application Layer Protocol
- T1573 Encrypted Channel
- T1008 Fallback Channels
- T1001 Data Obfuscation

## Sources

- [Threat landscape for SMBs in 2026: fake AI tools, phishing and more](https://securelist.com/smb-threat-report-2026/120357/)
- [Kaspersky SMB Threat Landscape 2026 - Full Report](https://securelist.com/smb-threat-landscape-2026/)
- [Trojware Campaign Targeting SMBs with Fake AI Tool Installers](https://www.kaspersky.com/enterprise/threats/trojware-ai-tools/)
- [Analysis of 1,100+ Malware Samples in AI Tool Impersonation Campaign](https://securelist.com/malware-analysis-ai-tools-trojware/)
- [Fake ChatGPT, Copilot, and Claude Installers Spread Trojware](https://www.kaspersky.com/blog/fake-ai-tool-trojware/)
- [How Attackers Are Impersonating AI Tools to Target SMBs in 2026](https://securityintelligence.com/ai-tool-trojware-smb-attacks/)
- [Phishing Email Campaign Analysis: Fake AI Tool Download Links](https://securelist.com/phishing-ai-tools-analysis/)
- [Trojware C2 Infrastructure Mapping: AI Tool Campaign](https://securelist.com/c2-infrastructure-fake-ai-tools/)

## Full data

Detection queries (Splunk SPL / Microsoft KQL / Sigma) and IOC values require the Threadlinqs MCP server (Purple tier): https://intel.threadlinqs.com/mcp

Canonical: https://intel.threadlinqs.com/threat/TL-2026-0993