DarkBERT: ChatGPT “devil friend” trained on Dark Web, should we be afraid of
DarkBERT, a specialized generativeAI model designed to understand darkweb data

DarkBERT: ChatGPT “devil friend” trained on Dark Web, should we be afraid of?

Beyond crafting our emails and social posts for work, generative AI has a broader scopem but what else can it do? Potentially, it could be CISO’s left hand man. 

Meet DarkBERT, a specialized generativeAI model designed to understand darkweb data. Pretrained with Dark Web data and armed against cybercrimes, It can deliver preemptive cybersecurity actionable advice, shielding entities and individuals from cyber threats like phishing and money laundering.

The dark web operates uniquely compared to the surface web we navigate daily. Dark web's access, structure, and language differ significantly. General AI models, like ChatGPT, lack the training and data to understand the dark web.

DarkBERT, as pertained with direct access to dark web specific data, grasps dark web terminology and logic adeptly. For instance, "Philipp Plein," a fashion designer on the surface web, signifies a drug type on the dark web. It can be used to detect ransomware leak sites, where hackers publish confidential data after failing to get a ransom, among many others. 

While DarkBERT holds potential in countering internet-based or internet-facilitated illegal activities, it is also being used by hackers as the foundation to build AI models for crimes. Its dual nature raises questions: will AI help combat the dark web, or inadvertently aid cybercriminals?

S2W, the Korean data intelligence company behind DarkBERT, will address these questions at Revive Tech Asia 2023.

Grab Your Free Ticket for Revive Tech Asia

 

DarkBERT ChatGPT Generative AI Cybercrime Hacker Threat Intelligence cybersecurity Revive Tech Asia S2W Creator of DarkBERT Speaking at Revive Tech Asia Tech event Tech conference 18 – 19 Oct sky100 Hong Kong CISO Ransomware data leak

 

 

 

Tags