AI Red Team Specialist — English & Portuguese (Global, Remote)
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 network of AI safety experts to stress-test large language models before they reach the public. As a Red Team Specialist, you'll act as an adversarial tester — probing AI systems with creative, challenging inputs designed to surface weaknesses, biased outputs, unsafe responses, and edge-case failures. Your findings directly shape the safety guardrails and training data that make AI systems more robust and trustworthy.
This is a flexible, remote-first opportunity for detail-oriented bilingual professionals who enjoy critical thinking, linguistic nuance, and the challenge of "breaking" AI systems in a controlled, constructive way.
What you'll do
- Design and execute adversarial prompts intended to expose model vulnerabilities, safety gaps, and unwanted behaviors
- Evaluate AI-generated responses for factual accuracy, tone, bias, and policy compliance in both English and Portuguese (global variant)
- Document and categorize discovered issues with clear, reproducible reasoning
- Collaborate with internal teams to refine red-teaming methodologies and reporting standards
- Provide structured feedback that feeds into model fine-tuning and safety training pipelines
- Maintain strict confidentiality and quality standards across all evaluation work
Requirements
- Native or near-native fluency in **English** and **Portuguese** (global Portuguese; Brazilian Portuguese speakers not eligible for this specific track)
- Strong analytical and critical thinking skills, with an eye for subtle inconsistencies or unsafe content
- Comfort working independently in a remote, asynchronous environment
- Familiarity with AI/LLM outputs, prompt engineering, or content moderation is a plus
- Excellent written communication skills for documenting findings clearly
- Reliable internet access and ability to commit consistent hours weekly
Compensation
Hourly rate: **$29–$45/hour**, based on experience and evaluation performance. This is a remote, contract-based engagement open to qualified candidates globally.
What we're looking for
- AI red-teaming
- Adversarial testing
- Bilingual content evaluation (English/Portuguese)
- Prompt engineering
- Content moderation
- AI safety review
- Critical analysis
- Technical documentation