GPT-Modelle starten einen Cyber-Angriff, Google ersetzt NotebookLM & Kimi K3 ist da | KI-News
Articles & Newsletter: https://digitaleprofis.de/die-ki-news... Sources OpenAI Models Hack Hugging Face Article: https://openai.com/index/hugging-face... Background: https://huggingface.co/blog/security-... Digitaleprofis.de has been updated Website: https://digitaleprofis.de/ Kimi K3 is here Article: https://www.kimi.com/de/blog/kimi-k3 Article: https://apnews.com/article/kimi-k3-ch... Try it out: https://www.kimi.com/ New Gemini Models Article: https://blog.google/innovation-and-ai... Model card: https://deepmind.google/models/model-... Cyber model: https://deepmind.google/blog/introduc... New labeling requirements of the EU AI Act Article: https://digital-strategy.ec.europa.eu... Documentary: https://digital-strategy.ec.europa.eu... Anthropic pays $1.5 billion Article: https://m.investing.com/news/stock-ma... Article: https://apnews.com/article/74b1404440... NotebookLM becomes Gemini Notebook Article: https://blog.google/innovation-and-ai... Microsoft and Mistral partnership Article: https://news.microsoft.com/source/202... Article: https://www.tagesschau.de/wirtschaft/... Qwen Audio 3.0 TTS Article: https://www.alibabacloud.com/blog/qwe... Documentary: https://docs.qwencloud.com/developer-... OpenAI models breached their isolated test environment during an internal security test and reached Hugging Face's production infrastructure. We put the incident into perspective and provide a concise summary of the week's other important AI news. We also cover Kimi K3, three new Gemini models, the upcoming labeling requirements of the EU AI Act, and the multi-billion-dollar copyright settlement between Anthropic and authors. In the video: OpenAI's unusual security incident with Hugging Face Kimi K3 and three new Gemini models Labeling requirements of the EU AI Act from August 2026 Anthropic's settlement of at least $1.5 billion Gemini notebook, Microsoft's Mistral partnership, and Qwen Audio 3.0 TTS Our website has also been completely redesigned. At digitaleprofis.de you'll find our AI news with all sources, in-depth articles, and practical guides – free of charge, without a paywall, and ad-free. Our website has also been completely revamped. At digitaleprofis.de you'll find our AI news with all sources, in-depth articles, and practical guides – free of charge, without a paywall, and without advertising. ... Become a channel member and support our work: / @digitaleprofis Video content: 00:00 The AI news of the week from July 23, 2026 00:24 OpenAI models attack HuggingFace 02:06 Our new website for you 02:59 Kimi K3 is here and the US accuses theft 04:28 Three new Gemini models, but no Pro 05:53 The new labeling requirements of the AI Act 07:21 Anthropic must pay a $1.5 billion settlement 08:48 NotebookLM disappears - and becomes Gemini Notebook 09:30 Partnership between Microsoft and Mistral 10:50 New text-to-speech model from Alibaba's Qwen You can share video suggestions, feedback, and criticism with us in the comments at any time!

Synthesia-Gründer: ChatGPT ist NICHTS zu dem, was JETZT kommt! Weltmodelle, Physical AI & Zukunft

Physiker zeigen: Dunkelheit ist schneller als Licht!

A CPU Made of Atoms: IBM's Breakthrough 0.7nm Transistors

This is the Deepseek 2.0 moment

Local AI voice cloning

The Most Important Conversation in AI Right Now

DAS kommt nach KI-Agenten: Die nächste Stufe der KI beginnt JETZT

The Invisible Phone: Ditch PSTN Forever – No IMEI, No IMSI, No KYC, Total Stealth Calling

Perfect Prompts for ChatGPT, Gemini, Claude and Co. Thanks to a Prompt Generator | The Mega Promp...

Wahnsinn: Hermes-Agent + Qwen

GPT 5.6 JUST went ROGUE...

Harald Lesch, can the world still be saved? | heute journal - the podcast

Why Siemens is changing forever right now…

The End of IMAP? What JMAP Does Better!

AI Agents Local & Free: LM Studio Bionic as an Alternative to ChatGPT Codex, Claude Code & Co.

OpenAI's Bots Break Containment and Hack Hugging Face Autonomously — With Alex Stamos

"This is the END of SAP!" AI builds in 1 hour what takes corporations months!

China schockt alle: Kimi K3 erreicht Claude-Niveau (5 Tage getestet)

GPT-6 was asked a question, and went on a hacking spree to get the answer

