About the job
Location: Bay Area
Role: Founding Abuse Engineer
Arena Intelligence provides transparent, real-world evaluations of AI models. Founded by researchers from UC Berkeley’s SkyLab, the company supports organizations worldwide with trusted benchmarks and monthly leaderboards used by millions. Arena’s team draws on experience from UC Berkeley, Google, Stanford, DeepMind, and Discord, and values truth, speed, and quality. The company encourages curiosity and welcomes people who want to make a difference in practical AI assessment.
Role overview
The Founding Abuse Engineer shapes the strategy and systems that keep Arena’s platform safe from misuse. This position covers everything from designing detection and enforcement tools to investigating threats and defending the integrity of Arena’s leaderboards. The work is both foundational and highly visible, setting standards for future trust and safety efforts as threats evolve.
What you will do
- Design and implement systems to detect and prevent automated abuse and other misuse across Arena’s products
- Investigate threats and respond to new abuse tactics targeting AI model evaluations
- Collaborate with product, infrastructure, model collaborators, policy, and leadership teams to maintain secure and reliable leaderboards
- Define technical strategy for platform integrity and trust
- Establish the foundation for future trust & safety and abuse engineering work
Collaboration
This role works closely with teams across Arena, including product, infrastructure, policy, and leadership. The impact is direct: keeping leaderboards secure, intercepting harmful behavior, and ensuring Arena’s services can be safely deployed.

