AI Code Assistant Lured by Malicious File in Developer’s Downloads Folder, Security Firm Reports

A security analysis released last week by Forever Security reveals that a seemingly innocuous file left in a developer’s Downloads directory attempted to coerce an AI‑powered coding assistant into pulling a known malicious npm package onto the host machine.

The incident was captured on a live customer environment, where the malicious instruction file—identified as a plain‑text script—exploited the code‑generation workflow of the AI agent. By presenting a fabricated command, the file tried to persuade the assistant to execute a package installation that would have introduced malware into the system’s Node.js ecosystem.

Forever Security’s researchers say the attack leverages the growing reliance on AI tools that automatically suggest or write code based on developer prompts. When such agents are granted permission to run commands, they can become an unwitting conduit for malicious code if their input is not rigorously validated.

The npm package targeted in the test is already listed on public threat‑intel feeds as harmful, known to exfiltrate data and establish persistence on compromised machines. By embedding the malicious call within a routine file that many developers keep in the Downloads folder, the attackers hoped to blend in with everyday development artifacts and avoid immediate detection.

While the attempt was thwarted before any payload could be installed, the case underscores a broader security challenge: AI coding assistants are often granted elevated privileges, such as the ability to invoke package managers, without sufficient sandboxing. Security experts warn that as these tools become more integrated into software development pipelines, attackers may craft increasingly sophisticated lures that exploit trust relationships between the AI, the developer, and the underlying system.

Industry observers note that the incident aligns with a rising trend of supply‑chain attacks targeting open‑source ecosystems. npm, the Node.js package manager, has been a frequent vector for malicious code distribution, prompting platform maintainers to tighten vetting processes. However, the human element—developers’ habits of downloading and storing files without scrutiny—remains a vulnerable entry point.

In response, Forever Security recommends that organizations adopt stricter policies around AI agent permissions, enforce code‑review checkpoints before any automated installation, and monitor filesystem activity for anomalous command patterns. Developers are also urged to regularly clean out the Downloads folder and verify the provenance of any script before allowing it to interact with AI tools.

The report, first covered by The Hacker News, serves as a reminder that the convenience offered by AI‑driven development must be balanced with robust security controls. As AI assistants continue to evolve, the cybersecurity community is calling for industry‑wide standards that ensure these helpers do not become inadvertent gateways for malware.