CVE & Application Security AI Task Auditor - Freelance AI Trainer Project
Pubblicato il 09-09-2026 - Jobgether in Italia
This position is listed on behalf of a partner company, who manages all applications and next steps. Our partner is looking for a CVE & Application Security AI Task Auditor - Freelance AI Trainer Project based in Italy.
This freelance opportunity is designed for experienced application security professionals who want to apply their expertise to the development and evaluation of advanced AI systems.
You will audit security-focused tasks used to train and evaluate AI models, ensuring they are technically accurate, realistic, reproducible, and practical.
The role involves reviewing Common Vulnerabilities and Exposures (CVE) scenarios, vulnerability assessments, secure coding practices, and application security workflows.
You will evaluate whether security tasks can be reliably solved, tested, and validated against clearly defined evaluation criteria.
Your technical judgment will help identify vulnerabilities, exploit execution flaws, mitigation issues, and other security-related inconsistencies.
The project offers a flexible,
remote environment where specialized security knowledge can directly influence the quality and reliability of AI evaluation workflows.
As a freelance contributor, you will work independently on challenging technical assignments while helping ensure AI systems are assessed against realistic application security scenarios.
Accountabilities
- Evaluate CVE and application security tasks to determine whether they are technically accurate, realistic, solvable, reproducible, and supported by reliable tests and evaluation criteria.
- Review security scenarios involving vulnerability assessment, secure coding, threat modeling, and application security practices for technical accuracy and practical relevance.
- Assess vulnerability descriptions, exploitation scenarios, expected outcomes, and mitigation approaches to ensure they are logically sound and technically valid.
- Identify security vulnerabilities, exploit execution flaws, mitigation logic errors,