We are hiring an AI Researcher to join the research team building the next generation of AI-native security systems. You will work alongside security and threat researchers to build large-scale AI agents that reason over software, code, endpoint activity, and security signals to detect malicious behavior, uncover vulnerabilities, assess risk, and make autonomous security decisions in real-world production environments. We are entering the Mythos era - where attackers operate at machine speed using autonomous systems and AI-generated software, and defenders must evolve the same way. We use state-of-the-art frontier models, including access to Mythos, to build reliable AI-native security systems at global scale. You will help design the evaluations, harnesses, and reliability infrastructure that make autonomous agents dependable under real customer load, while collaborating with leading AI organizations including Anthropic on initiatives such as Glasswing. This is an opportunity to work at the frontier of AI, autonomous systems, and cybersecurity while helping define how the next generation of security systems will operate.
Key Responsibilities
Build AI agents and autonomous security systems that reason over software, code, endpoint activity, MCPs, and security signals to detect malicious behavior, uncover vulnerabilities, and assess risk at production scale.
Develop systems, tooling, and infrastructure that enable agents to autonomously investigate threats, hunt for malware in massive datasets, and operate reliably in complex security environments.
Design and run experiments to evaluate frontier-model and agent capabilities in realistic adversarial scenarios, including benchmark creation, large-scale datasets, automated evaluations, and human-in-the-loop review systems.
Build the evaluation harnesses, observability systems, and reliability infrastructure required to make autonomous agents accurate, scalable, and dependable under real customer load.
Engineer for scale and performance across large distributed AI systems, including inference optimization, orchestration, batching, caching, cost controls, and graceful degradation under high demand.
Continuously evaluate emerging models, agent architectures, prompting techniques, and research directions to ensure our systems remain at the frontier of AI-native cybersecurity.
Rapidlyזprototype and test new approaches across reasoning, autonomy, evaluations, and security workflows as the AI landscape evolves.
Partner closely with threat and security researchers to extract domain expertise, translate analyst reasoning into AI workflows, and enable new forms of automation and autonomous investigation.
Collaborate with leading AI and security researchers to shape the future of AI-native cybersecurity as the industry transitions into the Mythos era.
Senior candidates will help define research direction, shape technical strategy, identify high-leverage problems, and influence how autonomous AI systems are deployed across the organization.
Requirements: Strong experience building and operating AI agents or autonomous systems in production environments.
Hands-on experience with LLMs, agent frameworks, tool use, reasoning systems, retrieval, evaluations, or multi-agent orchestration.
Proven ability to rapidly design experiments, iterate on ideas, and turn research into reliable production systems.
Deep familiarity with the rapidly evolving AI ecosystem; enthusiasm for continuously experimenting with new models, techniques, architectures, and research directions.
Strong intuition for identifying which new AI capabilities are production-ready versus hype, and ability to quickly translate frontier advances into practical systems.
Strong engineering skills, especially in Python and modern AI infrastructure.
Proven ability to own problems end-to-end, from research and prototyping through deployment, scaling, and reliability.
This position is open to all candidates.