AI Safety Red Teamer — English & Norwegian Bilingual Specialist
Paid remote expert bounty in Expert network (Global). Work in the software you already know; rates and ladder published up front.
- $30-$250
- per hr, published up front
- 1
- slots left of 1
- Remote
- no AI experience required
What you'll be paid to show
About the role
We're building a global red team of language experts who help make AI systems safer by finding their weaknesses before anyone else does. As an AI Safety Expert fluent in both English and Norwegian, you'll probe large language models with adversarial prompts, surface vulnerabilities, and generate the structured feedback that trains safer, more robust AI systems.
This is sensitive, high-impact work: you'll be reviewing model outputs that touch on bias, misinformation, harmful behaviors, and other edge cases where AI can go wrong. Your judgment and linguistic precision directly shape how safely these systems behave for real users.
What you'll do
- Design and execute adversarial prompts intended to surface unsafe, biased, or otherwise problematic model behavior
- Evaluate AI-generated outputs in both English and Norwegian for safety, accuracy, and appropriateness
- Document and categorize vulnerabilities, edge cases, and failure modes with clear, structured write-ups
- Collaborate with a distributed team of reviewers to calibrate standards and maintain consistent quality
- Provide detailed rationale and annotations that will be used to fine-tune and harden AI models
- Flag emerging risk patterns across sensitive topics such as misinformation, harmful content, and bias
Requirements
- Native or near-native fluency in both English and Norwegian (written and spoken)
- Strong analytical and critical thinking skills, with comfort engaging with sensitive or challenging content
- Excellent written communication for documenting findings clearly and precisely
- Ability to work independently in a fully remote setting with minimal oversight
- Prior experience with content moderation, linguistics, red-teaming, QA, or AI evaluation is a plus
- Reliable internet connection and availability for consistent, ongoing project work
Compensation
Hourly rate: **$48–$62/hour**, based on experience and evaluation performance. This is a flexible, remote contract engagement with opportunities for continued work based on project needs.
What we're looking for
- AI red-teaming
- Content evaluation
- Bilingual linguistic analysis
- Bias and misinformation detection
- Adversarial prompt design
- Data annotation and documentation