💼 About This Role

You'll design and run experiments to improve oversight of increasingly capable AI models as part of the Alignment Oversight team at OpenAI. You'll combine longer-horizon research with hands-on deployment, building practical oversight systems used today. This role offers the chance to produce externally publishable research and directly shape future model behavior.

🎯 What You'll Do

• Design and implement alignment experiments for oversight systems
• Deploy systems for action monitoring, red-teaming, and human-in-the-loop control
• Develop evaluations for alignment failure modes of frontier models
• Analyze deployment data to understand model failures and oversight gaps

📋 Requirements

• Hands-on experience training or evaluating large ML models, especially LLMs
• Experience with reinforcement learning or post-training methods
• Strong engineering execution to turn research ideas into reliable systems
• Ability to move quickly from research intuition to working experiments

✨ Nice to Have

• Experience with scalable oversight or model evaluation
• Familiarity with human-in-the-loop control systems
• Published research in alignment or AI safety

🎁 Benefits & Perks

• 💰 Competitive salary and equity
• 🏖️ Hybrid work model (3 days in office per week)
• 🚚 Relocation assistance
• 🏥 Comprehensive benefits (implied by equal opportunity employer)

📨 Hiring Process

Estimated timeline: 2-4 weeks · AI estimate

1. 1Recruiter call· 30 min
2. 2Technical interview· 60 min
3. 3On-site interviews· Full day