Path: csiph.com!fu-berlin.de!uni-berlin.de!individual.net!not-for-mail
From: rbowman <bowman@montana.com>
Newsgroups: comp.os.linux.misc
Subject: Re: Interesting - Ukraine Missiles Use Open Source, Russians Use Pi-4s
Date: 23 Jul 2026 04:14:21 GMT
Lines: 19
Message-ID: <ncdiotFuaclU10@mid.individual.net>
References: <nc6hn7F167tU2@mid.individual.net> <nc6ubeF3b9oU1@mid.individual.net> <113lhdo$1aeq8$1@dont-email.me> <nc77psF4pmvU1@mid.individual.net> <nc7bqlF5c1rU1@mid.individual.net> <113m1rg$1g23p$1@dont-email.me> <113mrla$1mo7o$4@dont-email.me> <113nej2$1rvch$4@dont-email.me> <nc9s4rF7pt3U11@mid.individual.net> <113oi8b$29i10$2@dont-email.me> <nca48rFjc9vU1@mid.individual.net> <ncadrjF7pt3U13@mid.individual.net> <113qapd$2s428$3@dont-email.me> <Ly48S.463810$khyb.63955@fx44.iad> <ncc4g0Fssh0U1@mid.individual.net> <nccgr7FuaclU7@mid.individual.net> <H8ydnYpicMk_EPz3nZ2dnZfqnPWdnZ2d@giganews.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
X-Trace: individual.net I0KecVGc/6n0KEDlB7GgvgUxZgGKmcWgpv1uFvMUYI2ao7V/Gy
Cancel-Lock: sha1:nt1y1SPKAJ1Pv5C/Ic2usNrnyow= sha256:F9a0xW9AJ4++dKn/yV8Ws4ka2ENoAo3AS4lNRLposrU=
User-Agent: Pan/0.165 (Kostiantynivka)
Xref: csiph.com comp.os.linux.misc:89001

On Wed, 22 Jul 2026 23:41:26 -0400, c186282 wrote:

>    News yesterday - AI system being tested deliberately BROKE OUT of its
>    "security cage" - cracked and infiltrated another company's database
>    in pursuit of it's "mission".

https://www.theregister.com/ai-and-ml/2026/07/23/openai-scored-an-own-
goal-with-huggingface-attack-showing-how-open-chinese-models-are-winning/
5276699

"The company's AI-culpa fits the narrative spun by US rival Anthropic 
about its Mythos models, which it deemed too dangerous to release except 
to totally trustworthy corporations and governments."

I'd put Anthropic at the head of the list of untrustworthy corporations 
and governments followed closely by every corporation and government on 
the planet.