Security Weaknesses Identified in AI Browsers
Summary
Hide ▲
Show ▼
Security researchers at SquareX Labs have identified architectural security weaknesses in AI browsers, including Perplexity’s Comet. AI browsers integrate AI assistants to automate user tasks, but this integration introduces new cyber risks. These risks include malicious workflows, prompt injection, malicious downloads, and misuse of trusted apps. The research highlights the need for stronger safeguards as AI capabilities become standard in web browsing.
Timeline
-
09.10.2025 16:15 1 articles · 4h ago
Security Weaknesses Identified in AI Browsers
Researchers at SquareX Labs have identified several architectural security weaknesses in AI browsers, including Perplexity’s Comet. These weaknesses include malicious workflows, prompt injection, malicious downloads, and misuse of trusted apps. The research emphasizes the need for stronger safeguards as AI capabilities become standard in web browsing.
Show sources
- Researchers Warn of Security Gaps in AI Browsers — www.infosecurity-magazine.com — 09.10.2025 16:15
Information Snippets
-
AI browsers integrate AI assistants to automate user tasks, such as searching, summarizing, and performing online actions through natural-language prompts.
First reported: 09.10.2025 16:151 source, 1 articleShow sources
- Researchers Warn of Security Gaps in AI Browsers — www.infosecurity-magazine.com — 09.10.2025 16:15
-
AI browsers from Perplexity, OpenAI, The Browser Company, and Fellou AI have been identified as having security weaknesses.
First reported: 09.10.2025 16:151 source, 1 articleShow sources
- Researchers Warn of Security Gaps in AI Browsers — www.infosecurity-magazine.com — 09.10.2025 16:15
-
Major platforms like Chrome and Edge are planning to add AI-driven capabilities.
First reported: 09.10.2025 16:151 source, 1 articleShow sources
- Researchers Warn of Security Gaps in AI Browsers — www.infosecurity-magazine.com — 09.10.2025 16:15
-
Security issues in AI browsers include malicious workflows, prompt injection, malicious downloads, and misuse of trusted apps.
First reported: 09.10.2025 16:151 source, 1 articleShow sources
- Researchers Warn of Security Gaps in AI Browsers — www.infosecurity-magazine.com — 09.10.2025 16:15
-
Existing security tools like SASE and EDR have limited visibility into AI browser behavior.
First reported: 09.10.2025 16:151 source, 1 articleShow sources
- Researchers Warn of Security Gaps in AI Browsers — www.infosecurity-magazine.com — 09.10.2025 16:15
-
Recommendations include establishing agentic identity systems, implementing DLP policies, adding client-side file scanning, and conducting extension risk assessments.
First reported: 09.10.2025 16:151 source, 1 articleShow sources
- Researchers Warn of Security Gaps in AI Browsers — www.infosecurity-magazine.com — 09.10.2025 16:15