Anthropic

Anthropic Seeks Cybersecurity Researchers to Advance N-Day Exploit System

🌐 Remote ⏱ Full-time 🎯 Mid Level
Apply Now
Anthropic Seeks Cybersecurity Researchers To Advance N-Day Exploit System
🏢
Company
Anthropic
📍
Location
San Francisco, CA, USA
Job Type
Full-time
🎯
Experience
Mid Level

About the Role

A leading AI research organization is seeking cybersecurity researchers to contribute to Mythos Preview, an advanced system capable of automating exploit development for N‑day vulnerabilities—those already disclosed and patched but still exploitable on unpatched systems. This role involves investigating newly reported software flaws and streamlining the creation of working exploits, a process that previously took days or weeks and can now be condensed to mere hours. It is a position for professionals who want to merge deep security expertise with cutting‑edge artificial intelligence to address some of the most urgent threats in modern infrastructure.

Key Responsibilities

  • Analyze newly disclosed N‑day vulnerabilities across operating systems, applications, and network services to assess exploitability.
  • Develop and test proof‑of‑concept exploits, leveraging both manual reverse‑engineering skills and the Mythos Preview automation framework.
  • Contribute to the improvement of the underlying AI models by providing curated exploit data and feedback on generated outputs.
  • Collaborate with AI researchers to refine machine learning techniques for vulnerability analysis and exploit generation.
  • Stay current with vulnerability disclosure channels, patch analysis, and emerging attack surfaces.
  • Document research findings and contribute to internal knowledge bases on exploit techniques and mitigation strategies.
  • Engage with the broader security community through responsible disclosure and publication of relevant research.

Requirements

  • Proven experience in cybersecurity research, particularly in vulnerability analysis and exploit development.
  • Proficiency in reverse‑engineering tools (IDA Pro, Ghidra) and debugging environments (WinDbg, GDB).
  • Strong programming skills in Python and C/C++ for building and automating exploit scripts.
  • Familiarity with modern exploit mitigation techniques and their bypasses (DEP, ASLR, CFG).
  • Understanding of machine learning concepts or prior work with AI‑augmented security tooling is a significant advantage.
  • Excellent written and verbal communication skills, with the ability to present technical findings clearly.

Compensation & Benefits

  • Competitive salary along with equity participation and performance‑based incentives.
  • Comprehensive health, dental, and vision insurance plans.
  • Flexible work arrangements, including remote and hybrid options.
  • Dedicated budget for professional development, conferences, and certifications.
  • Access to state‑of‑the‑art computing resources for security research.
  • Collaborative culture that values open research and responsible disclosure.

How to Apply

Interested candidates can submit their application through the Apply Now button on this page, which redirects to the official job listing. Please refer to the original posting for complete application instructions and any additional documentation requirements.

As N‑day exploitation continues to narrow the window between patch release and active attacks, this role sits at a critical intersection. For those with a similar focus on defensive cyber strategies, recent listings like DINUM Seeks Cybersecurity Incident Responders After Tchap Platform Hijack and Rockwell Automation Seeks OT Cybersecurity Experts for SecureOT AI Expansion also explore high‑impact security roles in rapidly evolving environments.

📨 Ready to Apply?

Apply before this opportunity closes.

Apply Now

📋 Job Overview

Company Anthropic
Location San Francisco, CA, USA
Job Type Full-time
Experience Mid Level
Posted Jun 9, 2026
Read Time 3 min read

✨ Interested in this role?

Apply Now