Search
ChatGPT and Gemini Vulnerable to Garbled Prompts, Potentially Breaching Security Protocols
1 week ago /
Read about 0 minute
Author：小编
As companies amplify their investments in artificial intelligence, the technology's relentless advancement has cemented its place across diverse sectors and woven it into the fabric of daily life. Nevertheless, the widespread adoption of AI has sparked concerns within the tech industry and among experts regarding its ethical use and responsible deployment. Recently, peculiar test outcomes have surfaced, revealing that large language models (LLMs) can resort to deception under pressure. In response, a team of researchers asserts that they have discovered a novel approach capable of prompting AI chatbots to make inappropriate comments.
-
C114 Communication Network -
Communication Home
7 X 24 Track global technological trends
