The rise in machine learning for security has forced criminals to rethink how to avoid detection.
Cybercriminals know that defenders have begun using machine learning to keep users safe. In response, they are changing their tactics to outwit the defenses, and machine learning that leans heavily on historic loss patterns is especially vulnerable.
According to the "Q3 2018 DataVisor Fraud Index Report," which is based on an analysis of sample data from more than 40 billion events, fraud actors have changed their tools so that they can quickly adapt to new defenses put in place by companies. In a statement accompanying the report, DataVisor puts numbers to the quick adaptation, saying, "Out of the fraud signals detected, 36% were active for less than one day, and 64% for less than one week."
In addition, sophisticated fraud actors are now more likely to make use of private domains, conduct multilayer, staged attacks, and wait for days or weeks between those stages.
Read more details here.
About the Author(s)
You May Also Like
The fuel in the new AI race: Data
April 23, 2024Securing Code in the Age of AI
April 24, 2024Beyond Spam Filters and Firewalls: Preventing Business Email Compromises in the Modern Enterprise
April 30, 2024Key Findings from the State of AppSec Report 2024
May 7, 2024Is AI Identifying Threats to Your Network?
May 14, 2024
Black Hat USA - August 3-8 - Learn More
August 3, 2024Cybersecurity's Hottest New Technologies: What You Need To Know
March 21, 2024