Great news!

[boplaid@mas.to]

Great news! Debian has a new GR to ban LLMs: https://www.debian.org/vote/2026/vote_002

@davidgerard

[mooooooo@qaf.men]

@boplaid @davidgerard Who gets to vote on this? Can users of Debian do something to protect their OS from slop or is this top-down?

[paul@notnull.space]

@boplaid @davidgerard
"It is our belief that widespread LLM usage comes from the "move fast, and break things" attitude that, while common in many parts of this industry, is contrary to what makes Debian Debian"

chef's kiss

This is definitely one to watch!

[womble@infosec.exchange]

@boplaid @davidgerard the counter-proposal being a tired re-hashing of the "ethical, legal, and environmental issues aside" talking points is a perfect distillation of the problem. The plagiarism machine is fundamentally incapable of ever producing anything original, and those who delegate their production to the machine quickly discover they can no longer produce anything original, either.

[phf@dmv.community]

@boplaid I guess we'll have to wait and see. Because I fear the "AI heads" will stink the place up something fierce and whoever is allowed to vote will worry about another fork and just keep their heads down. Sadly I have NO faith in Debian doing "the right thing" here. See systemd. @davidgerard

[zer0unplanned@friendica.rogueproject.org]

@boplaid @davidgerard Is that why Parrot OS uses AI in some tools?
officially introduced a dedicated AI tools category in its major 7.0 release (codenamed "Echo"), which launched in late 2025 based on Debian 13
The distribution's approach to AI is distinct: rather than using artificial intelligence to automate attacks, the new tools focus on securing AI systems and testing Large Language Models (LLMs) against emerging threats. The flagship addition is hexstrike-ai, a tool specifically designed for LLM prompt security, prompt engineering tests, and identifying vulnerabilities in AI-driven applications. This aligns with the project's 2026 roadmap, which prioritizes studying AI agents and models as new attack surfaces.
cybersecuritynews.com/parrot-7…
parrotsec.org/blog/2026-06-30-…

[cccpresser@chaos.social]

@boplaid @davidgerard

p.s. the proposal author @werdahias is on fedi

[dasgrueneblatt@wien.rocks]

@boplaid 🥹 @davidgerard