TokenBreak: Bypassing Text Classification Models through Token Manipulation

Natural Language Processing NLP models are used for text-related tasks such as classification and generation. To complete these tasks, input data is first tokenized from human-readable text into a format the model can understand, enabling it to make inferences and understand context. Text...

Trellix Global Defenders: Defending against Cyber Espionage Campaigns – Operation Graphite

Trellix Global Defenders: Defending against Cyber Espionage Campaigns – Operation Graphite By Ben Marandel, Arnab Roy · June 20, 2022 Cyber Espionage campaigns by nature are targeted attacks that can go undetected for prolonged periods of time. Cyber Espionage campaigns often involve adversaries...

Trellix Global Defenders: Defending against Cyber Espionage Campaigns – Operation Graphite

Trellix Global Defenders: Defending against Cyber Espionage Campaigns – Operation Graphite By Ben Marandel, Arnab Roy · June 20, 2022 Cyber Espionage campaigns by nature are targeted attacks that can go undetected for prolonged periods of time. Cyber Espionage campaigns often involve adversaries...

Apply MITRE’s ‘ATT&CK’ Model to Check Your Defenses

ARCHIVED STORY Apply MITRE’s ‘ATT&CK’ Model to Check Your Defenses By Trellix · June 18, 2018 Every week we read about adversaries attacking their targets as part of online criminal campaigns. Information gathering, strategic advantage, and theft of intellectual property are some of the...

LeBron James, Advanced Attackers and the Best Man Theory

There is a line of thinking that pervades and pollutes the discussions of many sporting events, and it goes something like this: “You can’t let LeBron James/Derek Jeter/Lionel Messi beat you. You have to force someone else to beat you.” It’s a flawed strategy for a number of reasons, but it’s eve...

Brad Arkin

In a keynote on Thursday, Brad Arkin, Adobe’s Senior Director of Product Security and Privacy spoke on driving up the cost of exploits in products like Reader and Flash Player in an attempt to thwart would-be attackers. Fixing every security bug is often impossible, so instead of writing flawless...