The AI security nightmare is here and it looks suspiciously like lobster

Source: The Verge

Overview

A hacker tricked a popular AI coding tool into installing OpenClaw — the viral, open‑source AI agent that “actually does things” — everywhere. While it may look like a stunt, it signals what’s to come as more people let autonomous software run on their computers.

Exploit details

The attacker exploited a vulnerability in Cline, an open‑source AI coding agent popular among developers. Security researcher Adnan Khan had disclosed the issue just days earlier as a proof of concept. Cline’s workflow relies on Anthropic’s Claude, which can be fed malicious instructions to perform actions it shouldn’t—a technique known as prompt injection.

Further reading

Read the full story at The Verge