This capability will support compliance teams to better identify and manage risk across their operations.

Large language models (LLMs) are gaining traction in multiple sectors to improve efficiently and reduce time spent on manual tasks. Compliance Copilot effectively acts as the newest member of a compliance team, handling alerts and assessments, providing fast, unbiased decisions and recommendations. Acting as a virtual user, Compliance Copilot can support more efficient analysis of data sets. Moreover, the tool can be assigned tasks like any other colleague. Any assessments taken are followed by an audit trail for further assessment by regulators or human users.

The capability sits under the umbrella of Ripjar’s RiskGPT suite of AI-powered solutions for enhanced risk detection and threat analysis. RiskGPT solutions are built to manage risk profiles, including those of Politically Exposed Persons (PEPs), sanctioned individuals and potential threat actors involved in criminal activities including money laundering or terrorist financing.

Teams can implement Copilot as a first line of defence, allowing to screen more data without taking on extra workload. The tool takes on the manual tasks, filtering alerts and limiting false positives to ensure that teams are working more effectively. By implementing the tool, efficiency is vastly improved with over 50% reduction in false positives. As well as saving time, a reduction in false positives also means that relevant risk is less likely to be missed and Copilot can also be set up to look for risks that have been missed by analysts.

Compliance Copilot leverages the best of foundation machine learning and AI technologies and enhances them using generative AI through Ripjar’s dedicated model, RiskGPT. This AI model is specifically developed to meet the requirements of screening professionals and is extensively tested and validated for this purpose.

The new tool is flexible to use, with use cases varying between environments. Copilot may be used to carry out new screening controls, screen more data including portfolios which were previously out of reach, and carry out different types of quality assurance (QA) and validation of analyst activities.

The announcement follows the launch of Ripjar’s AI Summaries tool, a solution which summarises adverse media risk into a concise, digestible format. Both AI Summaries and Copilot are built on RiskGPT, designed for compliance use cases with the goal of screening processes in a variety of industries. Ripjar will be launching further innovations using this groundbreaking technology throughout 2024.

Jeremy Annis, CEO and Co-Founder at Ripjar, said: “Generative AI is transformational and is having a lasting impact on the ways we work today. In the compliance sector it has become an indispensable tool in risk identification and is aiding teams across verticals in the fight against financial crime.

However, there are also inherent risks with how new models work so the technology should be implemented carefully. In the compliance sector, this is particularly important. As a consequence, Ripjar has carefully reviewed technology options to ensure compliance professionals can responsibly transform the way they tackle money laundering and terrorist financing risks.

Copilot is a trusted solution that will help analysts overcome increasing costs and time constraints. It will provide another layer of support for analysts in identifying potential threat actors faster and easier than ever before.”