Generative & Responsible AI - LLMs Use in Differential Governance
Atif Farid Mhammad, Ram Agarwal, Tony Columbo, Jess Vigorito · 2023
Generative AI techniques are widely used to extract valuable insights from data, which can be used to expand into new markets or save money by changing existing structures. However, there is a lack of innovation in the compliance area of financial transactional platforms, which can lead to money laundering. Standard Generative AI algorithms, specifically the use of Responsible AI within the realm of LLMs, have the potential to build robust detection systems to apply differential governance to capture money laundering transactions. Additionally, refined AI predictions can be used to improve the response time of compliance users in deciding whether a transaction is suspicious. This paper focuses on using Generative AI algorithms, anomaly detection, and network analysis to detect suspicious transactions and improve the overall risk rating assessment using hypothetical dataset. In this research we have proposed differential governance, oversight, controls and procedures incorporating generative AI, starting with differential privacy assisting financial institutions in detecting fraud while still upholding customer privacy rights and regulatory obligations. But it requires deliberate effort to implement it responsibly.