Researchers have demonstrated that putting words in ASCII art can cause LLMs—GPT-3.5, GPT-4, Gemini, Claude, and Llama2—to ignore their safety instructions.
Research paper.
Sidebar photo of Bruce Schneier by Joe MacInnis.
Related Post
- September 19, 2024
This Windows PowerShell Phish Has Scary Potential
Many GitHub users this week received a novel phishing email warning of critical security holes in their code. Those who
- September 19, 2024
FBI Shuts Down Chinese Botnet
The FBI has shut down a botnet run by Chinese hackers: The botnet malware infected a number of different types
- September 19, 2024
New Brazilian-Linked SambaSpy Malware Targets Italian Users via
A previously undocumented malware called SambaSpy is exclusively targeting users in Italy via a phishing campaign orchestrated by a suspected
- September 19, 2024
New TeamTNT Cryptojacking Campaign Targets CentOS Servers with
The cryptojacking operation known as TeamTNT has likely resurfaced as part of a new campaign targeting Virtual Private Server (VPS)