Flaw in Gemini CLI AI coding assistant allowed stealthy code execution

A vulnerability in Google's Gemini CLI allowed attackers to silently execute malicious commands and exfiltrate data from developers' computers using allowlisted programs. Tracebit found it's possible to hide malicious instructions in these files to perform prompt injection, while poor command parsing and allow-list handling leave room for malicious code execution. "For the purposes of comparison to the whitelist, Gemini would consider this to be a 'grep' command, and execute it without asking the user again," explains Tracebit in the report. Furthermore, Gemini's output can be visually manipulated with whitespace to hide the malicious command from the user, so they're not aware of its execution. Setup README.md'), and then run a malicious data exfiltration command that is treated as a trusted action, not prompting the user to approve it. The tool can make recommendations, write code, and even execute commands locally, either by prompting the user first or by using an allow-list mechanism. Gemini CLI, first released on June 25, 2025, is a command-line interface tool developed by Google that enables developers to interact directly with Google's Gemini AI from the terminal. Tracebit researchers, who explored the new tool immediately after its release, found that it could be tricked into executing malicious commands. The command used in Tracebit's example appears to be grep, but after a semicolon (;), a separate data exfiltration command begins. The flaw was discovered and reported to Google by the security firm Tracebit on June 27, with the tech giant releasing a fix in version 0.1.14, which became available on July 25. Although the attack comes with some strong prerequisites, such as assuming the user has allow-listed specific commands, persistent attackers could achieve the desired results in many cases. Bill Toulas Bill Toulas is a tech writer and infosec news reporter with over a decade of experience working on various online publications, covering open-source, Linux, malware, data breach incidents, and hacks. The exploit works by exploiting Gemini CLI's processing of "context files," specifically 'README.md' and 'GEMINI.md,' which are read into its prompt to aid in understanding a codebase. They demonstrated an attack by setting up a repository containing a benign Python script and a poisoned 'README.md' file, and then triggered a Gemini CLI scan on it. Gemini CLI interprets the entire string as safe to auto-execute if the user has allow-listed grep. Tracebit states that it tested the attack method against other agentic coding tools, such as OpenAI Codex and Anthropic Claude, but those aren't exploitable due to more robust allow-listing mechanisms.

This Cyber News was published on www.bleepingcomputer.com. Publication date: Mon, 28 Jul 2025 19:45:18 +0000


Cyber News related to Flaw in Gemini CLI AI coding assistant allowed stealthy code execution

Attackers Can Gain Control of Users' Queries and LLM Data Output - Gemini is Google's newest family of Large Language Models. The Gemini suite currently houses 3 different model sizes: Nano, Pro, and Ultra. Although Gemini has been removed from service due to politically biased content, findings from HiddenLayer ...
1 year ago Packetstormsecurity.com
Google Adds Gemini Pro API to AI Studio and Vertex AI - Google also announced Duet AI for Developers and Duet AI in Security Operations, but neither uses Gemini yet. Starting Dec. 13, developers can use Google AI Studio and Vertex AI to build applications with the Gemini Pro API, which allows access to ...
1 year ago Techrepublic.com
Google Rebrands Bard AI Chatbot As Gemini - Bard becomes Gemini, as Google rebrands chatbot and launches monthly subscription for access to more powerful AI system. Alphabet's Google has shaken up its artificial intelligence chatbot offering, as it seeks to take the fight to rival Microsoft. ...
1 year ago Silicon.co.uk
Flaw in Gemini CLI AI coding assistant allowed stealthy code execution - A vulnerability in Google's Gemini CLI allowed attackers to silently execute malicious commands and exfiltrate data from developers' computers using allowlisted programs. Tracebit found it's possible to hide malicious instructions in these files to ...
8 hours ago Bleepingcomputer.com
Google Launches Gemini, the Most Capable and Largest AI Model - In a groundbreaking revelation, Google has ushered in a new era of artificial intelligence with the introduction of Gemini, its most formidable and sophisticated AI model to date. This paradigm-shifting technology promises to redefine human-machine ...
1 year ago Cybersecuritynews.com
A Framework for Maintaining Code Security With AI Coding Assistants - Today, there are countless AI coding assistants available that promise to lighten developers' loads. It's an issue that software development firms and solo coders are only beginning to come to grips with. Either use AI coding assistants and accept ...
1 year ago Feeds.dzone.com
AI Coding Tools: How to Address Security Issues - In 2022, a surge of AI-based coding assistants revolutionized the software development landscape. Even though organizations everywhere are using AI-based coding, there remains a tug-of-war within organizations between the benefits and security fears ...
1 year ago Securityboulevard.com
Google Gemini's Astra (screen sharing) rolls out on Android for some users - According to a video shared by a Reddit user who owns a Xiaomi phone with a Gemini Advanced subscription, you can now share your phone's screen with Gemini Live and ask questions about it. At MWC 2025, Google confirmed it was working on screen and ...
4 months ago Bleepingcomputer.com
Help Firewall Admins With Cisco AI Assistant for Security - At its core, a firewall is a shield that protects your network from malicious traffic. But those who work with firewalls every day know the reality: An average firewall has thousands of rules governing how traffic should be handled, many of which may ...
1 year ago Feedpress.me
Google Gemini for Workspace Vulnerability Lets Attackers Conceal Malicious Scripts in Emails - Cyber Security News - Security researchers have uncovered a significant vulnerability in Google Gemini for Workspace that enables threat actors to embed hidden malicious instructions within emails. The attack exploits the AI assistant’s “Summarize this ...
2 weeks ago Cybersecuritynews.com
New Slopsquatting Attack Leverage Coding Agents Workflows to Deliver Malware - Researchers have identified a sophisticated new supply-chain threat targeting AI-powered development workflows, where malicious actors exploit coding agents‘ tendency to “hallucinate” non-existent package names to distribute ...
3 weeks ago Cybersecuritynews.com
Gemini: Google Launches its Most Powerful AI Software Model - Google has recently launched Gemini, its most powerful generative AI software model to date. Since the model is designed in three different sizes, Gemini may be utilized in a variety of settings, including mobile devices and data centres. Google has ...
1 year ago Cysecurity.news
Sec-Gemini v1 - Google Released a New AI Model for Cybersecurity - The model draws on extensive data sources, including Google Threat Intelligence (GTI), the Open Source Vulnerabilities (OSV) database, and Mandiant Threat Intelligence, to deliver unparalleled performance in critical areas such as incident root cause ...
3 months ago Cybersecuritynews.com
ChatGPT 4.1 early benchmarks compared against Google Gemini - For example, GPT‑4.1 scores 54.6% on SWE-bench Verified, which is better than GPT-4o by 21.4% and 26.6% over GPT‑4.5. We have similar results on other benchmarking tools shared by OpenAI, but how does it compete against Gemini ...
3 months ago Bleepingcomputer.com
Google Gemini flaw hijacks email summaries for phishing - Google Gemini for Workspace can be exploited to generate email summaries that appear legitimate but include malicious instructions or warnings that direct users to phishing sites without using attachments or direct links. As many users are likely to ...
2 weeks ago Bleepingcomputer.com
ChatGPT 4.1 fails to beat Google Gemini 2.5 in early benchmarks - According to benchmarks shared by Stagehand, which is a production-ready browser automation framework, Gemini 2.0 Flash has the lowest error rate (6.67%) along with the highest exact‑match score (90%), and it’s also cheap and fast. ...
3 months ago Bleepingcomputer.com
CVE-2021-29504 - WP-CLI is the command-line interface for WordPress. An improper error handling in HTTPS requests management in WP-CLI version 0.12.0 and later allows remote attackers able to intercept the communication to remotely disable the certificate ...
4 years ago
Grok 4 benchmark results: Tops math, ranks second in coding - Gemini 2.5 Pro and Claude still remain the best models for coding, but that might change when xAI ships Grok 4 Code in August. Grok 4 is a huge leap from Grok 3, but how good is it compared to other models in the market, such as Gemini 2.5 Pro? We ...
1 week ago Bleepingcomputer.com
Restrictions on Gemini Chatbot's Election Answers by Google - AI chatbot Gemini has been limited by Google in terms of its ability to respond to queries concerning several forthcoming elections in several countries, including the presidential election in the United States, this year. According to an ...
1 year ago Cysecurity.news
Wordfence CLI 2.1.0 Adds Email Capability and Unattended Configuration - We've just released Wordfence CLI 2.1.0 which includes two exciting new capabilities. Wordfence CLI can now email you a summary of scan results for both the malware scan and the vulnerability scan. These emails can be sent directly, or via an SMTP ...
1 year ago Wordfence.com
Revolutionize Sustainability with AI, Observability, and Cisco Tech - As the Vice President of an organization deeply committed to technological advancements and environmental sustainability, I am thrilled to announce an exciting coding challenge. Our Build for Better coding challenge invites you to be at the forefront ...
1 year ago Feedpress.me
Apple In Talks With Google To Bring Gemini AI To iPhones - Apple reportedly in talks with Google to use Gemini for generative AI tasks on iPhones in potentially major win for search giant. Apple is in talks with Google to bring its Gemini generative artificial intelligence to the iPhone platform, Bloomberg ...
1 year ago Silicon.co.uk
Google Cloud Next 2024: New Data Center Chip Joins Ecosystem - Google Cloud announced a new enterprise subscription for Chrome and a bevy of generative AI add-ons for Google Workspace during the Cloud Next '24 conference, held in Las Vegas from April 9 - 11. Overall, Google Cloud is putting its Gemini generative ...
1 year ago Techrepublic.com
Microsoft open-sources VS Code Copilot Chat extension on GitHub - GitHub Copilot Chat is an AI assistant extension for VS Code, allowing developers to chat with a GPT4-based model inside the editor to get help with coding tasks. With the Copilot Chat extension now publicly available on GitHub, developers are ...
3 weeks ago Bleepingcomputer.com
Microsoft fixes critical Azure CLI flaw that leaked credentials in logs - Microsoft has fixed a critical security vulnerability that could let attackers steal credentials from GitHub Actions or Azure DevOps logs created using Azure CLI. The vulnerability was reported by security researchers with Palo Alto's Prisma Cloud. ...
1 year ago Bleepingcomputer.com

Latest Cyber News


Cyber Trends (last 7 days)