Behind the Mask: Benchmarking Camouflaged Jailbreaks in Large Language Models

Large Language Models LLMs are increasingly vulnerable to a sophisticated form of adversarial prompting known as camouflaged jailbreaking. This method embeds malicious intent within seemingly benign language to evade existing safety mechanisms. Unlike overt attacks, these subtle prompts exploit...

Malwarebytes CrackMe 2: try another challenge

Last November, we released the first edition of the Malwarebytes CrackMe. Encouraged by the positive response we received from the security community, we decided to repeat the game, hopefully making it even more interesting and entertaining. As before, the CrackMe is dedicated to malware analysts...

Exploit for Out-of-bounds Read in Openssl

OpenSSL Heartbleed CVE-2014-0160 vulnerability scanner, dat...

The Pros and Cons of Letting the Kids Join Facebook

Following its $16 billion infusion from its recent IPO, the world’s largest social network is reportedly developing a technology that would allow children to access Facebook under parental supervision. A Facebook spokesperson neither confirmed nor denied that they are developing such technologies...