Remote | Freelance | Upcoming Project

Are you fluent in Dutch (NL)?

We’re currently looking for detail-oriented AI Claims Reviewers to join an upcoming project.

You’ll help evaluate AI-generated claims by comparing them against reliable sources and determining whether they are supported, partially supported, unsupported, or require further evidence.

Responsibilities:

• Review AI-generated claims in Dutch (NL) and flag uncertainty, gaps, inaccuracies, or overconfidence.
• Research and compare multiple sources, including news, official statements, databases, and other reliable references.
• Assess source credibility and determine whether the evidence supports, partially supports, or contradicts a claim.
• Produce structured verdicts: supported, partially supported, unsupported, or needs more evidence.
• Document reasoning clearly so other reviewers and users can follow your logic.
• Work within TrustScale's human-review workflow, which includes claims, sources, verdicts, and confidence scores.
• Identify factual inaccuracies, misleading statements, and unsupported conclusions in AI-generated content.
• Apply consistent research and fact-checking criteria across different topics and sources.
• Meet quality and turnaround standards for Dutch (NL))-language review queues.

Requirements:

• Strong proficiency in Dutch (NL) and English reading comprehension.
• Strong research, analytical, and critical-thinking skills.
• Ability to evaluate the credibility and reliability of online sources.
• Excellent attention to detail and ability to provide clear, evidence-based reasoning.
• Experience in fact-checking, journalism, research, content moderation, AI evaluation, or a related field is a plus.
• Ability to work independently and follow structured review guidelines.

About TrustScale:

TrustScale is an AI Trust and Verification Platform. TrustScale bridges the trust gap between AI and humanity through solutions that offer hallucination detection and reinforcement learning to reduce risk and prevent unintended consequences of AI outputs.