AI systems 'subject to new types of vulnerabilities,' British and US cyber agencies warn

Retrieved on: 2023-11-27 22:18:17

Tags for this article:

Cyberwarfare

Artificial intelligence

Cybernetics

Large language models

OpenAI

National Cyber Security Centre

AI safety

ChatGPT

Computer security

Generative pre-trained transformer

Regulation of artificial intelligence

Click the tags to see associated articles and topics

AI systems 'subject to new types of vulnerabilities,' British and US cyber agencies warn. View article details on HISWAI: https://therecord.media/ai-subject-to-new-vulnerabilities

Excerpt

The NCSC in August warned about “prompt injection attacks” as an apparently fundamental security flaw affecting large language models (LLMs) — the ...

Article found on: therecord.media

View Original Article