Google's Antigravity development tool for building AI Agents has been out for less than 11 days, and already the company has been forced to update its known issues pages after security researchers discovered what they believe are vulnerabilities. According to a blog post from Mindgard, one of the first to discover problems with Antigravity, Google is not classifying the issue it found as a security bug.
See also: Google starts showing ads in AI mode
However, Mindgard says a malicious actor could create a malicious rule by exploiting Google Antigravity’s strict guidance that any AI assistant it creates must always follow the rules set by the user. Author Aaron Portnoy, Mindgard’s head of research and innovation, says that after his blog was published, Google responded on November 25th saying that a report had been submitted to the product team responsible.
However, until action is taken, “the existence of this vulnerability means that users are at risk of backdoor attacks via compromised workspaces when using Antigravity, which can be exploited by attackers to execute arbitrary code on their systems. At this time, there is no setting that we can identify to protect against this vulnerability,” Portnoy wrote in his blog.
Even in the most restrictive mode, “the exploit proceeds uninterrupted and without user confirmation,” he wrote. When asked for comment, a Google spokesperson said the company is aware of the issue reported by Mindgard and is working to address it. In an email, Mindgard’s Portnoy said the nature of the bug makes it difficult to address.
See also: Google Gemini: Getting better at AI image detection
Other researchers have also found issues with Google Antigravity: Adam Swanda says he discovered and disclosed an indirect prompt injection vulnerability to Google. He said Google responded to him that the issue was a known issue and expected behavior of the tool. Another researcher using the name Wunderwuzzi wrote in a blog post about discovering five vulnerabilities, including data extraction and remote command execution via indirect prompt injection vulnerabilities.
This blog notes that, according to Google’s Antigravity known issues page, the company is working to resolve several issues. When asked for comment on the issues reported in the three blogs, a Google spokesperson said: “We take security issues very seriously and encourage reporting of all vulnerabilities so that we can quickly identify and address them. We will continue to publicly disclose known issues as we work to address them.”
See also: Google Maps: Gemini AI Tips, EV charger availability & anonymous reviews
What is Google Antigravity?
Antigravity is an integrated development environment (IDE), released on November 18, that leverages the Google Gemini 3 Pro chatbot. It is a development platform that combines a familiar, AI-powered coding experience with a new agent-first interface. This allows you to develop agents that autonomously design, execute, and verify complex tasks across your processor, terminal, and browser.
🔒 Protect your privacy with Proton VPN
Swiss VPN from the creators of Proton Mail — strict no-logs policy, strong encryption, and built-in NetShield that blocks ads, trackers, & malware.
- ✔ No-logs, based in Switzerland (except 14-Eyes)
- ✔ NetShield: blocks ads, trackers & malicious domains
- ✔ Covers all devices — free version available
The link is an affiliate link — SecNews may receive a commission at no additional cost to you. It does not affect the independence of our article writing.
