Ollama drama as 'easy-to-exploit' critical flaw found in open source AI server

A now-patched vulnerability in Ollama – a popular open source project for running LLMs – can lead to remote code execution, according to flaw finders who warned that upwards of 1,000 vulnerable instances remain exposed to the internet. Wiz Research disclosed the flaw, tracked as CVE-2024-37032 and dubbed Probllama, on May 5 …

[Anonymous Coward]

O-LOL-ma

Run LLMs, get what you deserve.

[AliceThrees]

Re: O-LOL-ma

I just got the cute ollama sticker. No one deserves to be hurt. Update your Ollama!

[MonkeyJuice]

Re: O-LOL-ma

If you're dumb enough to expose a server port to the world, you only have yourself to blame, frankly.

[Korev]

Whipping the Ollama's ass?

[AliceThrees]

whoop whoop. You should update your software more often. At least 11 versions behind to be vulnerable.

[Doctor Syntax]

"An attacker could exploit the flaw by sending a specially crafted HTTP request"

Little Bobby Models?

[AliceThrees]

I love Ollama! It seems crazy that the other posters forget that they should deploy tools in their own VPC or set-up authentication themselves.

[Doctor Syntax]

Are we to take it that your posts are Ollama output?

[Anonymous Coward]

Resident ollama ?