The OWASP Top 10 [1] is the probably the most well-known and recognised reference standard for the most critical web application security risks. This organisation has now started to working on creating a similar list for Large Language Models (LLM) Applications.
I'm posting about it here since I think it would be beneficial for safety alignment researchers to be involved for two reasons:
- To provide AI safety and alignment expertise to the security community and standardisation process.
- To learn from the cybersecurity community both about standardisation processes since they have a long experience in developing these kinds of standards and about security mindset and vulnerabilities.
I have no idea about how many people within the AI safety and alignment community actually know about this initiative, but I did not find any reference to it on the alignment forum or here on the EA forum, so I thought I might as well post about it.
More information available here:
- https://owasp.org/www-project-top-10-for-large-language-model-applications/
- https://github.com/OWASP/www-project-top-10-for-large-language-model-applications
- ^
More info about The Open Web Application Security Project here https://en.wikipedia.org/wiki/OWASP