Hackers disguise infostealer malware as Claude Code and OpenClaw to target developers

Malicious Campaign Exploits AI Developer Tools

Kaspersky researchers discovered in March 2026 a sophisticated malicious campaign targeting software developers searching for AI developer tools like Claude Code and OpenClaw1

1

. The attack leverages sponsored Google search results to direct users to fraudulent installation pages that closely mimic official documentation2

2

. When developers search for terms like "Claude Code download" or "OpenClaw download," malicious ads appear at the top of search results, redirecting to fake documentation pages hosted on Squarespace3

3

.

How Infostealer Malware Infiltrates Systems

The fake documentation pages are visually identical to legitimate installation instructions, making them extremely difficult to distinguish from authentic sources1

1

. The attack exploits a common developer practice of using copy and paste commands for installation. Instead of installing the intended AI coding assistant, the malicious commands deploy different infostealers based on the operating system. Windows users receive Amatera, an information-stealing malware that collects data from user directories, web browsers, and cryptocurrency wallets3

3

. Kaspersky noted that Amatera operates under a Malware-as-a-Service (MaaS) model and has been observed in previous campaigns using the ClickFix distribution technique1

1

. Meanwhile, macOS users are infected with AMOS, a notorious infostealer previously documented in multiple campaigns targeting Apple devices2

2

.

Risks of Exposing Sensitive Data to Corporate Networks

Vladimir Gursky, cybersecurity expert at Kaspersky, emphasized that this malicious campaign poses significant risks because AI developer tools are widely used by professional developers working in large organizations3

3

. Infected victims may unknowingly expose source code from active projects, confidential corporate data, authentication credentials, and private accounts1

1

. The campaign specifically targets developers who rely on AI-assisted coding tools, making it particularly dangerous for businesses. Beyond Claude Code and OpenClaw, Kaspersky identified similar attacks disguised as AI developer tools including Doubao, an AI tool from ByteDance2

2

.

Pattern of Malvertising Attacks Against AI Tools

This isn't the first time attackers have exploited interest in AI tools through malvertising. In December 2025, Kaspersky detected that attackers spread a macOS infostealer using Google Ads by creating a fake ChatGPT tutorial that guided users through installing the Atlas Browser3

3

. The malicious instructions appeared to be hosted on a legitimate site associated with OpenAI, helping attackers gain users' trust. This pattern suggests that as AI coding assistants like Claude Code gain popularity, attackers will continue to create fake documentation pages and sponsored ads to distribute infostealer malware.

Protecting Against Malware Disguised as Claude Code

Kaspersky recommends several protective measures for developers. Always verify download links and ensure they point to official project websites rather than clicking on sponsored ads2

2

. The malicious sites used in these attacks are hosted on Squarespace, which is visible in the URL and should raise suspicion. Review any command-line instructions before executing them, especially when copied from external sources, and avoid following guides you don't fully understand3

3

. Implementing reliable security solutions that can detect infostealers like Amatera and AMOS is critical for organizations whose developers use AI-assisted coding tools. As AI tools continue gaining traction among software developers, expect these attacks targeting software developers to intensify, particularly through fake installation pages and malicious commands designed for stealing sensitive data including credentials and cryptocurrency wallet information.