Forever publishes EMBRACE findings on embedded AI browsers

Forever Security published EMBRACE research on embedded browsers in AI assistants, covering prompt injection, address spoofing and local-file exposure. It reported fixes for Cursor and Copilot, with Codex and VS Code findings unresolved at publication.

What the public record establishes

Forever published the grouped research on October 29. Its Cursor demonstration required a user keystroke and depended on model behavior; it reported a patch on October 1 after a September report. Copilot’s address-spoofing issue was fixed independently in build 1.1.29 on October 13, before Forever sent a report. [2]

Forever reported local-file exposure in Codex and VS Code as unresolved at publication. Microsoft rejected the VS Code report under its Workspace Trust boundary; Forever said OpenAI had not replied to its private report. These were researcher demonstrations, not reported malicious exploitation. [2][3]

Sources

  1. EMBRACE: How We Hacked the Browsers Embedded in Top Five Desktop AI Assistants

    Forever Security · Oct 29, 2026

  2. EMBRACE: [Technical Overview] How We Hacked the Browsers Embedded in Top Five Desktop AI Assistants

    Forever Security · Oct 29, 2026

  3. Researchers broke the browsers built into VS Code, Cursor, Copilot, Codex and Claude Desktop

    Help Net Security · Oct 29, 2026