Trenchant
Chrome Sandbox Escape Android - Vulnerability Researcher
- Compensation
- Not disclosed
- Location
- Global
- Work setup
- Remote
- Seniority
- Senior
- Employment
- Not specified
- Experience
- Not specified
- Category
- Mobile
- Posted
Skills
Must have
- - Demonstrable delivery of Chrome/Chromium sandbox escapes, browser-process vulnerabilities or closely comparable cross-privilege exploitation.
- - Deep knowledge of Chromium’s multi-process architecture, sandbox policy and renderer-to-browser trust boundaries.
- - Strong practical experience with Mojo IPC, bindings, data pipes, shared memory, interface ownership and message validation.
- - Advanced C++ vulnerability-research and exploitation skills in complex multi-process targets.
- - Working knowledge of Android internals relevant to Chrome, including application isolation, SELinux domains, Binder and platform services.
- - The ability to turn a subtle boundary mistake into a stable result that can be integrated into a wider chain.
- - Independent research ownership and a consistent history of finishing high-difficulty work.
- Strong signals
- - Credited Chrome sandbox escapes or comparable real-world cross-privilege exploit delivery.
- - Custom Mojo fuzzers, IPC introspection tools or Chrome instrumentation frameworks.
- - Experience chaining renderer compromise through sandbox escape to Android system or kernel impact.
- - Research across multiple Android OEMs, chipsets and Chrome branches.
- - vulnerabilities found made it to stable/beta releases.
- - Strong patch-analysis and variant-hunting results.
- How we work
- - Fully remote, with high autonomy and close collaboration between browser, platform and kernel specialists.
- - We measure progress through technically meaningful, reproducible delivery.
- - Public CVEs are not a requirement.
Full description
We are looking for a senior researcher with deep practical experience identifying and exploiting vulnerabilities across Chrome sandbox boundaries on Android.
You should already know how to move from a compromised renderer or low-privilege browser process to a meaningful trust-boundary violation. The goal is stable, chainable sandbox-escape capability on real Android targets.
What you’ll work on
- Renderer-accessible Mojo interfaces, interface brokers and browser-process endpoints.
- GPU, media, network, utility, device and other services reachable from low-privilege Chrome processes.
- Android-specific Chrome integration, including platform bridges, Binder-facing components, permissions and service boundaries.
- Confused-deputy conditions, validation gaps, unsafe deserialisation, lifetime errors, races and state-machine mistakes.
- Cross-process exploitation where asynchronous IPC, handles, shared memory or capability transfer affect reliability.
- End-to-end chains with renderer and Android-kernel researchers when needed.
What you’ll deliver
- Original vulnerabilities that cross a Chrome process, privilege or trust boundary on Android.
- Reliable sandbox-escape exploit components that work under production mitigations.
- Prioritised attack-surface areas that are deemed to be complex enough to contain vulnerabilities.
- Triggers, PoCs, exploitability analysis, target assumptions and complete technical handover.
- Reusable tooling for IPC discovery, Mojo message generation, tracing, instrumentation, coverage and variant analysis.
What we’re looking for
- Demonstrable delivery of Chrome/Chromium sandbox escapes, browser-process vulnerabilities or closely comparable cross-privilege exploitation.
- Deep knowledge of Chromium’s multi-process architecture, sandbox policy and renderer-to-browser trust boundaries.
- Strong practical experience with Mojo IPC, bindings, data pipes, shared memory, interface ownership and message validation.
- Advanced C++ vulnerability-research and exploitation skills in complex multi-process targets.
- Working knowledge of Android internals relevant to Chrome, including application isolation, SELinux domains, Binder and platform services.
- The ability to turn a subtle boundary mistake into a stable result that can be integrated into a wider chain.
- Independent research ownership and a consistent history of finishing high-difficulty work.