Engineering Leader, AI Safety Engineering & Evaluation, Apple Services Engineering

Written by the T&S Network from Apple's posting.

Leads an emerging AI safety engineering and evaluation organization within Apple Services Engineering, covering AI-powered experiences across the App Store, Apple Music, and Apple TV. The team builds automated, model-based methods to identify, measure, and mitigate harmful content, bias, robustness, and Responsible AI risks, working with Evaluation Systems & Research on shared evaluation infrastructure.

• AI misuse and jailbreaks

What you'd do

• Set the multi-year technical roadmap for AI safety engineering and evaluation and align it with organizational goals.
• Build and mentor a team of engineering managers, engineers, ML engineers, and applied scientists, owning staffing and performance.
• Lead development of LLM-as-a-judge, multi-model, and automated classification approaches to assess harmful content and bias.
• Direct automated red teaming, synthetic data generation, and scenario-based testing to surface systematic failure patterns.
• Convert manual expert safety reviews into repeatable automated processes embedded in model release workflows.
• Partner with ESR to integrate safety-specific models and assessments into shared evaluation platforms.
• Translate safety and governance requirements from privacy, legal, and safety policy into measurable technical criteria.

What it asks for

• 12+ years in software engineering, ML engineering, applied AI, or related disciplines, with production system delivery.
• 8+ years in engineering or applied ML management, including 3+ years managing engineering managers or multiple teams.
• Track record of growing engineering teams and setting up their technical practices from the ground up.
• Practical knowledge of ML and modern AI architectures, including inference and evaluation methods.
• Strong Python and modern software development practices.
• Bachelor's degree in a technical field or equivalent practical experience.
• Experience leading teams building AI/ML evaluation systems or automated model assessment (preferred).

Terms in this role

Red teaming.

Testing a system, often an AI model, by deliberately trying to make it fail or misbehave, so problems are found and fixed first.

Our best summary of the posting, and it can be wrong. The employer's posting is what counts: read it in full on Applenorth_east before you apply. If we got something wrong, report this listing.