Resistant AI protects high-risk automated decisions taken by machine learning, AI, and statistical models in the fintech and financial sector. We secure the existing credit risk assessment, fraud detection and anti-money laundering systems. Any of these applications offers high reward to any adversary that can manipulate or evade them at scale.
Motivated adversaries can and do use modern AI techniques or simple cunning to commit fraud by exploiting the vulnerabilities of machine learning methods and inconsistency or insufficiency of the training data. Resistant AI provides:
* forgery detection (for documents including payslips, statements and other documents),
* continuous security monitoring to discover manipulation or attacks, and
* significant reduction of false alarms or rejections.