GitHub Copilot is changing how the world builds software—and protecting that experience from abuse is critical to our users, our customers, and the integrity of the platform. We're looking for a data analyst to focus on identifying, investigating, and reporting on abuse, misuse, and policy violations of GitHub Copilot and other AI-powered developer products.

In this role, you'll turn Copilot telemetry, account signals, and usage data into clear insights that inform detection, policy, and enforcement decisions. You'll partner closely with Trust & Safety, Product, Engineering, Security, and Legal (CELA) to surface emerging abuse trends and help shape how we respond to them.

• Detect and investigate Copilot abuse. Analyze telemetry, usage logs, billing/entitlement data, and account signals to identify abuse patterns such as automated or scripted usage, account and credential sharing, token misuse, evasion of usage limits, prompt injection, content-policy violations, and fraudulent or coordinated activity.

• Build the analyses and tooling that power Trust & Safety decisions. Develop dashboards, reports, and prototype models (e.g., abuse-risk scoring, anomaly detection, investigation views) that help stakeholders distinguish legitimate Copilot usage from abuse and measure the impact of mitigations.

• Tell the story behind the data. Synthesize complex analyses into clear, actionable insights—through dashboards, written reports, and stakeholder presentations—that influence detection rules, policy updates, and enforcement actions.

• Improve data quality and coverage. Identify gaps in abuse-relevant data sources, recommend new pipelines or integrations, and partner with Engineering and Data Science to strengthen the signals available for Copilot abuse detection.

• Anticipate analytical tradeoffs. Evaluate methods and assumptions critically, surface risks like false positives/negatives and unintended impact on legitimate users, and recommend mitigations.

• Operationalize what works. Turn one-off investigations into reusable, self-service reporting where it makes sense, and help scale recurring abuse-monitoring capabilities.

• Collaborate across the company. Work with Trust & Safety, Copilot Product, Anti-Abuse Engineering, Security, Data Science, and CELA to align on data definitions, abuse taxonomies, and shared standards.

• Handle sensitive data responsibly. Apply data privacy, governance, and ethical handling practices to Copilot prompt, completion, and user-activity data, and ensure work goes through appropriate CELA review.

• Mentor and influence. Share domain expertise with peers, coach less experienced analysts, and present findings to senior Copilot, Trust & Safety, and Security leadership.

Required Qualifications

• 5+ years experience in data analysis and reporting, data science, business intelligence, or business and financial analysis o OR Bachelor's Degree in Statistics, Mathematics, Analytics, Data Science, Engineering, Computer Science, Business, Economics or related field AND 3+ years experience in data analysis and reporting, data science, business intelligence, or business and financial analysis
  • o OR equivalent experience.

Preferred Qualifications

• 7+ years of relevant experience, or a Bachelor's/Master's degree in a related field with 5+/3+ years of experience respectively

• 1+ year of experience with SQL, Python, Power BI, or comparable data analytics tools

• 2+ years of consulting experience

• Experience analyzing abuse, fraud, or Trust & Safety signals on a large-scale online platform—ideally involving AI/ML products such as GitHub Copilot or similar AI developer tools

• Familiarity with common abuse patterns on developer platforms (automated/scripted usage, account and credential sharing, evasion of usage limits, prompt injection, content-policy violations)

• Experience partnering with Trust & Safety, Anti-Abuse, Security, and Legal teams on sensitive investigations

Skills You'll Use

• Analytics & Storytelling: Data analysis, data modeling, data visualization, data storytelling, statistics, quality reviews

• Trust & Safety: Abuse and anomaly detection, behavioral analytics and pattern recognition, investigative analysis, Trust & Safety analytics

• Technical: SQL, Python, big data, database querying, automation, data integration, data cleaning

• Collaboration & Influence: Technical communication, consulting, influencing without authority, cross-functional partnership

• Judgment: Decision-making, problem solving, operational excellence, compliance and data governance