How to weaponize LLMs to auto-hijack websites

Caute_cautim — Sun, 18 Feb 2024 19:34:51 GMT

Hi All

Do we put this under Threats or Tech Talk or even Privacy too it crosses a lot of discussions:

AI models, the subject of ongoing safety concerns about harmful and biased output, pose a risk beyond content emission. When wedded with tools that enable automated interaction with other systems, they can act on their own as malicious agents.

Computer scientists affiliated with the University of Illinois Urbana-Champaign (UIUC) have demonstrated this by weaponizing several large language models (LLMs) to compromise vulnerable websites without human guidance. Prior research suggests LLMs can be used, despite safety controls, to assist [PDF] with the creation of malware.

https://www.theregister.com/2024/02/17/ai_models_weaponized/

Regards

Caute_Cautim

Re: How to weaponize LLMs to auto-hijack websites

Kyaw_Myo_Oo — Mon, 26 Feb 2024 15:00:54 GMT

Thank you for sharing this information with us @Caute_cautim.

topic Re: How to weaponize LLMs to auto-hijack websites in Tech Talk

How to weaponize LLMs to auto-hijack websites

Re: How to weaponize LLMs to auto-hijack websites