Fraud Investigator for AI Model Training [Remote]

Join a cutting-edge initiative focused on building realistic enterprise environments for training and evaluating frontier AI agents. As a Financial Services Fraud Investigation Expert, you will leverage your extensive experience in fraud investigation and financial crime to recreate the digital workspaces of major financial institutions. Your expertise will help design tasks that challenge state-of-the-art AI, ensuring they are grounded in real-world scenarios.

• Construct a realistic digital workspace that reflects your daily operations, including case files, investigation write-ups, SAR/CTR drafts, typology memos, alert-triage notes, and relevant email threads, while incorporating the platforms you utilize (e.g., Refinitiv World-Check, Splunk SIEM).
• Develop multi-step tasks based on your actual workflows that require interaction with various applications, files, and stakeholders, effectively challenging advanced AI agents.
• Collaborate with fellow fraud and financial crime experts to design the environment, define task scope, and review scenarios for accuracy and rigor.
• Engage asynchronously with research teams to refine task designs and establish evaluation criteria for fraud-investigation agent benchmarks.
• Contribute to pioneering AI research and benchmarking, with your work directly influencing how leading labs train and evaluate the next generation of AI systems.

• Minimum of 3 years of full-time experience in fraud or financial crime at a Fortune 500 bank, large card issuer, fintech, or insurance firm.
• Experience in areas such as BSA/AML compliance, sanctions screening, fraud operations, financial-crime investigations, insurance claims fraud, or regulatory exam responses.
• Certifications such as CAMS, CFE, or CFCS are a plus.
• Proficient in using tools like Refinitiv World-Check, Splunk SIEM, and FIS/Fiserv.
• Strong analytical thinking and writing skills, with the ability to translate investigations into structured task specifications.

• Task Completion Pay: Competitive compensation based on task quality, ranging from $1, 150 to $1, 450 per completed task, subject to change as the project evolves.
• Performance Bonus: Top performers will receive a weekly bonus incentive in addition to their per task rate.
• Hourly Opportunity: High-performing individuals may be invited to transition to an hourly compensation model based on consistent quality and throughput.