From Noise to Clarity: Generative AI in Adverse Media Screening by Alex Aoun

Six years ago, when I broached the topic of Artificial Intelligence (AI) in discovery conversations with background check providers, it often proved challenging. Many exhibited either fear or distrust, sometimes mingled with genuine interest. This fear and distrust stemmed from a misunderstanding of AI, which sometimes took on an air of mysticism.

Then ChatGPT, with its Generative AI (Gen AI) capabilities, broke everything wide open. The subsequent excitement and exuberance about AI remain to this day, but is now tempered with a dose of caution—an example of this being the EU AI Act.

Certainly, the emergence of Gen AI brought AI front and center, but AI has been around for a very long time. Several subsets of AI exist today—Machine Learning (ML), a subset of AI that involves training algorithms on data so they can make predictions or decisions without explicit programming; Deep Learning, a specialized subset of ML that uses neural networks with many layers (hence “deep”) to analyze complex patterns in large datasets; Natural Language Processing (NLP)—and, of course, Gen AI. They are all intertwined in some way.

‍

Breakthrough in Natural Language Processing

The breakthrough in Natural Language Processing (NLP) set the stage for Gen AI. NLP combines computational linguistics with statistical, machine learning, and deep learning models. This combination allows machines to understand, interpret, and predict an output in a meaningful way, much like a human would.

‍

Read the full article here, or access the Journal if you are a PBSA member.