Published • loading... • Updated
Researchers ‘Get AI Drunk’ to Test Chatbot Cyber Risks, UNSW Study Finds
Summary by cybernoz.com
This story is only covered by news sources that have yet to be evaluated by the independent media monitoring agencies we use to assess the quality and reliability of news outlets on our platform. Learn more here.4 Articles
4 Articles
Science and Haitek: Researchers at the University of New South Wales have shown that language models that are forced to simulate the speech of a drunken person are becoming markedly less resistant to jailbreak attacks.
Australian Cyber Security MagazineReposted by 
cybernoz.com
cybernoz.comResearchers ‘get AI drunk’ to test chatbot cyber risks, UNSW study finds
Researchers at the University of New South Wales (UNSW) say large language models (LLMs) become more likely to leak confidential information and answer restricted prompts when they are trained or prompted to mimic “drunken” speech patterns.
Coverage Details
Total News Sources4
Leaning Left0Leaning Right0Center0Last UpdatedBias DistributionNo sources with tracked biases.
Bias Distribution
- There is no tracked Bias information for the sources covering this story.
Factuality
To view factuality data please Upgrade to Premium
