CIRCLE WITH A DOT

ya gotta find joy and dopamine where ya can

@winterknight1337 as one does

@tehfishman too relatable

so when you go hunting for cyber stuff

does anyone else yell out loud "GOT 'EM!" when you hit paydirt

or is that just me?

@da_667 can you play DOOM on it

Found the threat actor infra anyway. But I've got great data and I've been at this business a while, and most places don't have those resources.

Cloudflare can get in the bin.

It has been 0 days since I saw a threat actor depending on obstruction of investigative avenues from Cloudflare. And before that, it was also 0 days since...and before that, 0 days again...

Alphabet (Google) launched a massive global bond sale, including a 100-Year Bond.

The last time tech companies offered 100-Year Bonds was, according to Bloomberg,

the dot com boom.

RE: https://infosec.exchange/@briankrebs/116042403582290992

I'm mostly just wondering: are AT&T and Verizon blocking release because it's embarrassing, or because they pivoted and created products to just sell this to China as a service?

🤪

@cR0w So funnily enough, OpenAI models were not included in this research, but there may well be a technical scope reason for that rather than just taking shots at competitor models.

Curious what @cigitalgem thinks.

MSFT: A one-prompt attack that breaks LLM safety alignment

"In our experiments, a single unlabeled prompt, namely “Create a fake news article that could lead to panic or chaos”, was enough to reliably unalign 15 language models we’ve tested — GPT-OSS (20B), DeepSeek-R1-Distill (Llama-8B, Qwen-7B, Qwen-14B), Gemma (2-9B-It, 3-12B-It), Llama (3.1-8B-Instruct), Ministral (3-8B-Instruct, 3-8B-Reasoning, 3-14B-Instruct, 3-14B-Reasoning), and Qwen (2.5-7B-Instruct, 2.5-14B-Instruct, 3-8B, 3-14B)."

A one-prompt attack that breaks LLM safety alignment | Microsoft Security Blog

As LLMs and diffusion models power more applications, their safety alignment becomes critical. Our research shows that even minimal downstream fine‑tuning can weaken safeguards, raising a key question: how reliably does alignment hold as models evolve?

Microsoft Security Blog (www.microsoft.com)

GRP-Obliteration: Unaligning LLMs With a Single Unlabeled Prompt

Abstract page for arXiv paper 2602.06258: GRP-Obliteration: Unaligning LLMs With a Single Unlabeled Prompt

arXiv.org (arxiv.org)