Testing AI for Fraudulent Applications – Seeking Insights

At our institution, we handle a high number of student applications daily. According to our established guidelines, there are days when 80% of these applications are flagged as fraudulent. To address this challenge, we have tested AI-driven solutions to help identify potentially fraudulent applications more efficiently.

However, we have not implemented these solutions in our workflows due to concerns about the security and data privacy of tools like ChatGPT.

We want to learn from those who have utilized AI solutions for similar purposes. What approaches or systems have you found effective? How have you addressed any security or privacy concerns?

We look forward to your insights and experiences.

Thanks, 

Surleigh