• by Ray Schultz , June 26, 2020

Email security firm Ironscales has deployed natural-language processing (NLP) to detect common business email compromise (BEC) attacks.  

The firm uses machine learning and neural networks to help firms stem attacks that bypass legacy email gateway defenses, Ironscales says.  

This helps improves BEC protections while “relieving security analysts from the burden of manually analyzing and remediating every single threat,” states Eyal Benishti, founder and CEO of Ironscales. 

Clients seem to approve of the enhancement. “Not a day goes by that our executives aren’t impersonated,” states Amir Freudinger, IT manager at Nano Dimension. Freudinger adds that Ironscales has improved its protections “while not adding additional work for the security team.” 

According to Ironscales, the four most common types of BEC attack are:  

• Employee availability checks 
• Requests for an unspecific task
• Requests for a gift card 
• Requests to change direct deposit, bank details or request for payment

The firm notes an FBI report showing that BEC attacks cost businesses $1.77 billion in 2019. 

Bad actors increasingly focus on social engineering messages that do not contain malicious attachments or links, making them harder to detect, it adds.