Apple
AIML - Machine Learning Engineer, Evaluation - Responsible AI and Safety
$166K–$292K
Cambridge, Massachusetts · Full-time · Posted Oct 29, 2026
About the job
Evaluates and improves the safety and fairness of generative AI products by building evaluation methods, datasets, and pipelines that inform product decisions.
What you'll do
- Develop safety and fairness metrics, evaluation datasets, sampling strategies, and human- and automated-grading methods.
- Build and optimize data pipelines and automate workflows to deliver scalable evaluations before and after launch.
- Investigate failures, communicate findings, and provide actionable safety recommendations to product and policy teams.
What they ask for
- Develop evaluation metrics for generative-model safety and fairness risks.
- Design datasets and scale data coverage using human and synthetic generation methods.
- Develop sampling strategies that combine human annotation with auto-grading.
- Design and optimize data pipelines; work with highly sensitive content, including offensive and controversial material.
Nice to have
- Experience with multimodal models, agentic systems, or diverse languages and locales.
- Experience with AI compliance and regulation, large-scale benchmarking, or child safety, clinical/health, or companionship chatbot applications.
Pay
$165,800 and $292,400 USD base salary
Tools
large language models, diffusion models
AI focus
Responsible AI evaluation for generative AI, including safety, fairness, inclusivity, alignment, and monitoring of language, image, and multimodal models.
Model-written summary · the original posting is authoritative