Cognistx and Kroll
AI-Enabled Data Breach Detection

“To effectively monitor the massive amounts of personal and corporate information available on the deep and dark web requires the help of artificial intelligence technology,” says Sanjay Chopra, CEO of Cognistx, an applied artificial intelligence (AI) company. “AI reduces the need for manual review and helps make sense of this diverse data source; it synthesizes critical information automatically to add business value,” he explains.

Founded in 2015, Pittsburgh-based Cognistx creates adaptive, multi-strategy AI-powered computer systems that extract meaningful insights to produce results for clients. Chopra continues, “Cognistx’s strength lies in applying AI, natural language processing, and deep learning techniques to solve business problems.”

Since Cognistx’s inception, it has formed a strong partnership with Kroll, a world leader in risk management, investigations, and cyber resilience, to identify threats and reduce misuse of sensitive information leaked on the dark web. This partnership has contributed to the development of Kroll’s AI-powered CyberDetectER^® suite of solutions, for which it mines about 22 million disparate documents from the dark web on a daily basis.

For example, Kroll’s CyberDetectER BINWatch™ alerts financial institution clients if it detects their associated bank identification numbers (BINs) or credit card information on the dark web. “We are leveraging machine-learning capabilities to analyze files faster, gauge exposure levels, and mitigate risk in real time,” says Brian Lapidus, leader of Koll’s Identity Theft and Breach Notification business.

Cognistx utilizes AI clustering techniques, advanced junk modeling, and deep QA learning techniques on data, developed under the direct supervision of Dr. Eric Nyberg, Chief Data Scientist of Cognistx. Along with cutting-edge technologies, Cognistx boasts a team of leading researchers in the field of cognitive computing and machine intelligence to help achieve clients’ analytical objectives.

Cognistx’s technology and expertise help enable Kroll to identify relevant information quickly and minimize the time required for analysis.

Combined with Kroll’s extensive knowledge base that is continuously monitored by its global team of cyber experts and ability to quickly build a breach response plan, the companies are able to envision and develop sophisticated cyber solutions for ever-evolving data security threats.

“The partnership between Cognistx and Kroll has enabled both companies to enhance their skillsets and has led to successful outcomes, and sometimes eye-opening moments for clients,” states Lapidus. By way of example, Lapidus notes how a team of Cognistx and Kroll experts detected a breach that involved U.S. tax forms. They further discovered that these stolen documents were being traded for cash and other commodities on the dark web.

We are leveraging machine-learning capabilities to analyze files faster, gauge exposure levels, and mitigate risk in real time

A common challenge across industries is capturing the required data and normalizing it. For instance, Kroll needs to detect in real-time the data pertaining to client corporations and personnel that have been exposed on the dark web to mitigate fraud and other harmful activities. Cognistx reduces the number of files that need to be analyzed through its junk classification techniques; it also measures the results to determine the business value brought by AI.

Machine learning along with input from analysts is used to better train Cognistx’s detection models. Moving forward, Cognistx and Kroll plan to continue their collaboration as they look to develop unique AI-enabled cybersecurity solutions that will help organizations better prevent, detect, and mitigate the harms caused by data breaches.