Research Engineer – Human Influence

Responsibilities

We are representing a confidential AI safety research organization based in London. In this role, you will contribute to critical work examining how advanced AI systems can influence human beliefs, decisions, and behaviour—a substantial class of AI risk that operates through people.

Your responsibilities will include:

• Designing and building reinforcement learning environments aimed at mitigating concerning model behaviors, such as deception in conversational contexts or multi-agent interactions
• Applying state-of-the-art interpretability methods to identify why models exhibit concerning behaviors and designing mitigations that work across different training regimes
• Building scalable system architecture for repeatable delivery and analysis of model evaluations and benchmarks
• Delivering ambitious, engineering-heavy research projects using post-training techniques on large compute clusters
• Running rigorous human-AI interaction studies, randomized controlled trials, and developing datasets to monitor real-world risk exposure

What We're Looking For

Essential qualifications:

• Proven experience deploying a benchmark, evaluation, or product to users—such as an evaluations pipeline, open-source contribution, or similar large-scale research contributions
• Clear understanding of current AI safety literature and interest in human influence topics
• Experience fine-tuning or post-training LLMs using standard methods with libraries like PyTorch, Keras, JAX, or custom code
• Comfortable working with RL environments and using RL or reward-based methods to post-train or finetune ML models, ideally LLMs
• Writing scalable and maintainable production code in Python
• Experience serving, scaling, and containerizing ML code using tools like Docker, Kubernetes, Ray, FastAPI, or SLURM, especially on large compute clusters
• Clear understanding of fundamental machine learning concepts
• Clear and consistent communication skills

Desirable qualifications:

• Understanding of model internals for mechanistic interpretability research or analyzing model activations and weights
• Experience shipping an AI safety pipeline to production beyond research prototypes
• Experience using data to answer complex research questions by scoping, training, and validating classifiers or finetuned LLMs
• Experience with frontend and node.js deployments

How We Work with You

Get A Job.ai is representing this confidential organization in their search for exceptional talent. Here's how the process works:

• You apply through the Get A Job.ai platform
• Our talent team conducts an initial screening to ensure alignment with requirements
• We submit qualified candidates to our client for consideration
• Please do not attempt to contact the client directly—all communication flows through Get A Job.ai

This multidisciplinary team welcomes candidates from diverse backgrounds. You'll work on highly interdisciplinary projects drawing on computational social science, AI safety and security, cognitive and behavioral science, machine learning, and data science.

Pay

Pay via Get A Job.ai will be determined based on role scope and relevant experience. The typical range for this position is £65,000–£145,000 annually, which includes base salary plus technical allowance. An additional employer pension contribution of 28.97% is paid on the base salary component.

Benefits include hybrid working arrangements in London, generous annual leave (25+ days plus public holidays), paid parental leave, learning and development stipends, and comprehensive pension contributions.

Equal Employment Opportunity: Get A Job.ai is committed to fostering an inclusive recruitment process. We welcome applications from candidates of all backgrounds and circumstances.