Mozilla: ChatGPT Can Be Manipulated Using Hex Code LLMs tend to miss the forest for the trees, understanding specific instructions but not their broader context. Bad actors can take advantage of this myopia to get them to do malicious things, with a new prompt-injection technique.
Regulators Combat Deepfakes With Anti-Fraud Rules Despite the absence of laws specifically covering AI-based attacks, regulators can use existing rules around fraud and deceptive business practices.
LATEST FROM DR TECHNOLOGY
Sophos-SecureWorks Deal Focuses on Building Advanced MDR, XDR Platform Sophos CEO Joe Levy says the $859 million deal to acquire SecureWorks from majority owner Dell Technologies will put the Taegis platform — with network detection and response, vulnerability detection and response, and identity threat detection and response capabilities — at the core.
China's 'Evasive Panda' APT Debuts High-End Cloud Hijacking A professional-grade tool set, appropriately dubbed "CloudScout," is infiltrating cloud apps like Microsoft Outlook and Google Drive, targeting sensitive info for exfiltration.
AI Chatbots Ditch Guardrails After 'Deceptive Delight' Cocktail 
_Stephen_Barnes-Business_Alamy_Stock_Photo.jpg%20%20)
_Brain_light_Alamy.jpg%20%20)
_William_Mullins_Alamy.jpg%20%20)
