How to apply for Trusted Red Team

About Syntony

Syntony is a research and advisory firm that works with frontier AI labs, public institutions, and private organizations on safety research, governance design, adversarial evaluation, and risk advisory. It offers project-based work on confidential, multi-disciplinary engagements that directly inform client decision-making.

About the role

This is a project-based red team role. You will conduct scoped adversarial testing, threat modeling, and robustness evaluation for AI systems, frontier labs, and public sector clients. Your findings will become actionable evidence for governance decisions, mitigation strategies, and oversight controls.

A typical day

A typical day may involve scoped adversarial testing, threat modeling, or robustness evaluation for a client AI system. You might document findings into evidence for governance decisions or mitigation strategies. The work is confidential and multi-disciplinary, so you may coordinate with policy, security, and forecasting experts under NDA.

Who Syntony is looking for

• Has published work or demonstrated expertise in areas such as ML security, prompt engineering, policy analysis, geopolitical forecasting, or systems thinking.
• Can design and run scoped adversarial evaluations and threat models for AI systems under NDA.
• Is comfortable turning technical and policy findings into clear evidence for governance and oversight decisions.
• Prefers project-based, multi-disciplinary engagements over a single ongoing full-time role.

Tips for this application

• Highlight specific adversarial evaluation or red teaming projects you have done, including the scope, methods, and how findings were used.
• Show expertise across at least two of the listed areas: ML security, prompt engineering, policy analysis, geopolitical forecasting, systems thinking.
• Provide links to published work, public reports, or talks that demonstrate your judgment and technical depth.
• State clearly that you are open to project-based, NDA-bound work rather than a traditional full-time role.
• Explain how your findings have influenced governance, mitigation, or oversight decisions in past engagements.

What to cover in your cover letter

Emphasize your experience conducting scoped adversarial testing and threat modeling for AI systems. Describe how you have turned findings into actionable evidence for governance or mitigation. Point to published work or demonstrated expertise in relevant areas. Confirm your interest in project-based, confidential engagements under NDA.

Research before applying

• Read Syntony's website and any public reports to understand their safety research, governance design, and adversarial evaluation work.
• Look for public examples of Syntony's engagements with frontier AI labs, public institutions, or private organizations.
• Check the backgrounds of Syntony's team to see what disciplines and methods they prioritize.
• Find out how Syntony defines 'trusted red team' and what types of clients request this work.

Likely interview topics

Based on the job description, expect questions about:

• How you scope an adversarial evaluation for an AI system when the client's goals are confidential or unclear.
• A specific threat model you built for an ML system and what it revealed.
• How you translate technical red team findings into evidence that governance bodies can act on.
• Your approach to prompt engineering as an adversarial technique and its limits.
• How you handle multi-disciplinary work that combines policy analysis, geopolitical forecasting, and systems thinking.

Common mistakes to avoid

• Applying as if this is a standard full-time role instead of a project-based engagement.
• Claiming expertise in all listed areas without evidence of published work or demonstrated experience.
• Ignoring the NDA and confidential nature of the work in your application materials.