Hackaday reported on NLWeb, Microsoft’s new agentic web protocol that lets users integrate whatever LLM-based chatbot they desire
. However, what they didn’t do was protect themselves from a path traversal vulnerability which allowed any attacker who exploited it the means to steal keys and other sensitive information:
Although Microsoft patched it already, no CVE was assigned, while raising the question of just how many more elementary bugs like this may be lurking in the protocol and associated software […] Even aside from the the strong ‘solution in search of a problem’ vibe, it’s worrying that right from the outset it seems to introduce pretty serious security issues that suggest a lack of real testing […]
We can all talk about how this technology is constantly rushed out til we’re blue in the face, along with the reasons why. But it’s clear to me with cases like this that in that rush, they’ve decided it’s best to put it out with known and unknown vulnerabilities and patch as they go which is wild when you really think about it. A whole industry thriving on stealing resources from the planet and humans and a bunch of info sec people making money off bounties as they crop up like whack-a-moles. 10/10 system.
Filed under: Microsoft software