Telegram (AI) YouTube Facebook X
Ру
Vulnerability in Cursor AI Editor Allows PC Takeover

Vulnerability in Cursor AI Editor Allows PC Takeover

Vulnerability in Cursor AI editor allows covert PC takeover.

Opening a project folder in certain code editors can lead to the covert execution of malicious commands. According to SlowMist, users of Cursor AI are particularly vulnerable to this flaw.

The vulnerability affects popular development environments and tools for Vibe Coding, where programming is delegated to large language models.

The attack mechanism involves creating a project with a specific structure. If a developer opens such a folder using the standard Open Folder function, a malicious command is automatically executed on their device. The threat is relevant for both Windows and macOS.

According to experts, several users of the Cursor AI editor have already fallen victim to the campaign. The exact damage is unknown.

The founder of SlowMist, known by the pseudonym Cos, has already passed information about the incident to the platform’s security team.

At the time of writing, Cursor has not commented on the vulnerability reports.

Web3 researcher known as DeFi Teddy recommended users employ separate devices for Vibe Coding and cryptocurrency storage.

“Never open or download projects in Cursor from unverified or suspicious sources (such as random repositories on GitHub) whose security is not confirmed,” he added.

In September, Oasis Security specialists discovered a similar vulnerability in the program. It allowed malicious code to be embedded, taking control of the workspace and stealing API tokens without any user commands.

Cursor is an IDE based on Visual Studio Code with built-in AI tools. The project is integrated with popular chatbots like ChatGPT and Claude.

The platform is popular among developers: according to media reports, about a million people use it, generating over a billion lines of code daily. In May, the company behind Cursor, Anysphere, raised $900 million at a valuation of $9 billion.

Back in July, the cybersecurity service Tracebit found a vulnerability in Google’s Gemini. It allowed the stealthy execution of malicious commands if a user viewed suspicious code with the help of a neural network.

Подписывайтесь на ForkLog в социальных сетях

Telegram (основной канал) Facebook X
Нашли ошибку в тексте? Выделите ее и нажмите CTRL+ENTER

Рассылки ForkLog: держите руку на пульсе биткоин-индустрии!

We use cookies to improve the quality of our service.

By using this website, you agree to the Privacy policy.

OK