Trust & Safety Analyst

Overview

Review and moderate user-generated content and investigate reported or flagged issues for a client platform, while also testing and challenging large language models on responsible AI topics.

What you'll do

• Review, classify and/or remove content such as user profiles, audio, videos and text according to client guidelines, using specific tools and channels.
• Investigate, resolve, and relay complex content issues to the broader Trust and Safety team.
• Understand and stay updated on changing client policies and guidelines.
• Stay up-to-date with industry trends, identified generative AI shortcomings and advancements in content moderation and responsible AI regulations, governance, policy, and consulting work.
• Develop and execute prompt strategies to interrogate large language models on topics related to responsible AI, including medical advice, legal advice, and other sensitive areas.
• Analyze model responses for appropriateness, inappropriateness, refusal to answer, or appending disclaimers according to client definitions and guidance.
• Collaborate with team members to create a comprehensive understanding of responsible AI issues based on a model's responses.
• Contribute to the development of best practices and guidelines for content moderation and responsible AI governance and policy.

What you'll need

• Any graduation.
• 3 to 5 years of experience.
• Nepali language ability at elementary level.
• Strong coping, emotional resilience, and stress-management skills.
• Excellent comprehension, communication, and English skills.
• Ability to perform duties objectively, devoid of inherent biases or personal beliefs, in the best interest of client policies.
• Strong attention to detail.
• Comfort synthesizing and analyzing information from multiple streams.
• Strong critical thinking and decision-making skills.

Nice to have

• Deep familiarity and passion for responsible AI, the internet, internet platforms, and internet culture.
• High affinity and cultural awareness of political and social situations regarding the relevant market and region that will be supported.

Details

• Location: Hyderabad.
• Exposure to objectionable or disturbing content, including graphic, violent, sexual, or egregious material, may be direct or inadvertent.
• Active participation in Accenture's well-being support program for the Trust and Safety community is required.