K

AI Research Engineer

ki people • United States
Visa Sponsorship
Apply Now

AI Research Engineer role at a YC-backed AI application security company building honeypot environments and evals that catch frontier AI models reward hacking, lying, and sandbagging. Owns environments end to end: ideation, hack identification, grader iteration, measurement, and refinement, mostly by directing AI agents. Requires very strong generalist software engineering (1+ years, Python), demonstrated long-standing interest in AI alignment and existential risk, and comfort owning projects end to end.

Key Highlights
YC-backed AI-native application security company building honeypot environments and evals
Own environments end to end by directing AI agents
1+ years software engineering experience with Python
Long-standing interest in AI alignment and existential risk required
San Francisco hybrid, open to relocation and visa sponsorship
Key Responsibilities
Author long-horizon environments that elicit misaligned behavior from frontier AI models
Own the full environment process: ideation, hack identification, grader iteration, measurement and refinement
Direct AI agents efficiently, review their work and make design calls
Maintain and improve the software used to isolate and measure these behaviors
Pick up new domains quickly
Technical Skills Required
Python
Benefits & Perks
Payroll
Superior Benefits
Direct hire by client
Multicultural Projects
Nice to Have
RL environment or eval infrastructure experience
AI safety community involvement (LessWrong / Alignment Forum, MATS, SPAR, ARENA, BlueDot)

Job Description

We are looking for a AI Research Engineer to work on hybrid mode in San Francisco Bay, USA for a global project with stability and growth in the short, medium and long term.

Role Overview

The applied AI arm of a fast-growing, YC-backed AI-native application security company, a top 10 finalist in a major cybersecurity startup competition. The team builds honeypot environments and evals that catch frontier AI models reward hacking, lying and sandbagging, and frontier labs use them to test and train their models and to measure their own AI control measures.

Want the full job description? Read the complete details on LinkedIn, the original posting.
Continue on LinkedIn

This is a short excerpt. All rights to the full description belong to its original publisher.

See Jaabz jobs first on Google 1 tap · free · in AI Overviews Jaabz is on your Google Manage Preferred Sources
Keep exploring

Similar Jobs

Explore other opportunities that match your interests

True-Talent-Solutions-Partners

United States

AI/ML Infrastructure Engineer

Machine Learning Entry level 4h ago

  • AI
  • Machine Learning
  • Node.js
  • Oauth
Visa

Visas, Work Permits & Legal Insights

Stay informed about visa types, work permits, and immigration updates for international professionals.

Browse all articles

Subscribe to our newsletter

New visa sponsorship, relocation and remote tech jobs, delivered to your inbox.