Microsoft AI Monetization Platform Trust & Safety team seeks a mid-level professional with 3+ years experience in Trust & Safety, Risk Management, Fraud Prevention, or related fields. The role involves identifying vulnerabilities, designing adversarial testing, simulating attacker behaviors, creating datasets, and improving AI/ML model performance through collaboration with Engineering, Data Science, and other teams. Requires strong analytical skills, experience with AI/ML systems, cross-functional collaboration, and communication skills. Preferred experience includes adversarial testing, red-teaming, Generative AI, LLMs, and tools like Python, SQL, and Power BI. Location: Bengaluru, Karnataka, India.
What you'll do
Partner with Engineering, Data Science, Operations, Policy, Risk, and Security teams to proactively identify vulnerabilities, abuse vectors, and emerging threats across Trust & Safety systems
Design and execute adversarial testing programs that challenge detection models, enforcement systems, review processes, and operational workflows
Similar jobs
More roles worth a look
Related opportunities based on specialty and working model so candidates can keep momentum.
Bellevue, United States; Boston, United States; Boulder, United States; Chicago, United States; Denver, United States; Los Angeles, United States; New York, US; San Francisco, United States; San Jose, United States; Seattle, United States; Ventura, United States•7 days ago
Simulate real-world attacker behaviors, fraud tactics, prompt attacks, evasion techniques, policy circumvention attempts, and abuse patterns to uncover system weaknesses
Identify, document, and prioritize vulnerabilities, failure modes, blind spots, and emerging risks, driving mitigation strategies with partner teams
Create adversarial datasets, benchmark sets, attack corpora, and test scenarios to evaluate system robustness, reliability, and resilience
Perform hands-on labeling, content review, evaluation, and adjudication activities as needed to establish ground truth, characterize new attack patterns, validate vulnerabilities, and improve model performance
Generate actionable insights from investigations, attack simulations, labeling exercises, evaluation programs, and operational datasets to identify emerging trends and evolving threat patterns
Partner with Engineering and Data Science teams to improve model performance through adversarial testing, error analysis, red-team evaluations, feedback loops, and continuous learning mechanisms
Define red-teaming methodologies, attack taxonomies, threat models, and evaluation frameworks to systematically assess Trust & Safety defenses
Monitor external abuse trends, fraud ecosystems, industry developments, and emerging AI-powered attack techniques relevant to digital advertising and online platforms
Design experimentation and validation frameworks to assess the effectiveness of new defenses, controls, models, and enforcement mechanisms
Drive cross-functional initiatives from threat discovery through mitigation while balancing customer experience, operational efficiency, safety, and business objectives
Communicate technical findings, attack patterns, risk assessments, and strategic recommendations to technical and non-technical stakeholders
Embody Microsoft’s Culture and Values
Requirements
Bachelor's Degree in Business, Public Policy, Communications, Operations, Economics, Psychology, Social Sciences, Data Science, Analytics, Computer Science, Information Systems, Engineering, or a related field, or equivalent practical experience
3+ years of experience in Trust & Safety, Risk Management, Fraud Prevention, Abuse Prevention, Integrity Systems, Security, Threat Intelligence, Quality Operations, Product Management, Program Management, Analytics, Marketplace Integrity, or related domains
Experience identifying vulnerabilities, abuse patterns, operational risks, adversarial behaviors, quality gaps, or system weaknesses using data-driven approaches
Strong analytical and problem-solving skills with experience investigating complex issues, generating insights, and driving risk mitigation initiatives
Experience working across cross-functional teams including Engineering, Data Science, Operations, Policy, Risk, and Business stakeholders
Demonstrated ability to translate ambiguous threats, risks, or quality issues into actionable recommendations, evaluation frameworks, or operational improvements
Familiarity with AI/ML-powered systems and the role of adversarial testing, human review, labeling, evaluation, and feedback loops in improving model quality and safety outcomes
Experience evaluating products, models, workflows, or operational processes using experimentation, metrics, audits, data analysis, or structured investigations
Strong communication, stakeholder management, and influencing skills
Preferred: Familiarity with Generative AI, Agentic AI, Large Language Models (LLMs), jailbreak techniques, prompt attacks, model evasion techniques, and adversarial AI evaluation methodologies
Preferred: Experience generating actionable insights from large-scale investigation, evaluation, risk, quality, enforcement, or operational datasets and translating findings into model, product, or policy improvements
Preferred: Working knowledge of SQL, Power BI, Python, or similar analytics and reporting tools
Preferred: Experience partnering with Engineering and Data Science teams to improve AI/ML systems through red-teaming, evaluation design, labeling strategies, error analysis, experimentation, and feedback-loop mechanisms
Preferred: Demonstrated curiosity, investigative mindset, and ability to identify emerging risks before they become material business or ecosystem issues
Preferred: Proven ability to thrive in fast-paced, ambiguous environments while driving measurable customer, quality, safety, and business impact
Tech stack
AIMLPythonSQLPower BIGenerative AILarge Language ModelsAgentic AI
Apply now
Ready to take the next step in your career? Click the button below to continue to the application process.
