Tootfinder

Anthropic: KI könnte sich bald auf Haushaltsgeräten selbst replizieren
Anthropic warnt: Kleinere KI-Modelle könnten sich bald zum Beispiel auf Haushaltsgeräten selbst vervielfältigen. Kontrolle wäre dann kaum noch möglich.
…

Anthropic: KI könnte sich bald auf Haushaltsgeräten selbst replizieren
Anthropic warnt: Kleinere KI-Modelle könnten sich bald zum Beispiel auf Haushaltsgeräten selbst vervielfältigen. Kontrolle wäre dann kaum noch möglich.

Risk report: Anthropic raises misalignment risk estimate from very low to low and says it doesn't plan to release a stronger internal model called "Model 2" (Madison Mills/Axios)
https://www.axios.com/2026/08/14/anthropic-model-2-ai-risk

Anthropic sees AI risks rising, no plan to release stronger "Model 2"
The company is forging ahead even amid concerns about models' growing cybersecurity risks.

«KI-Zugänge werden zur Ware: Wie gestohlene Session-Tokens einen Schwarzmarkt für Frontier-Modelle antreiben»
So intelligent sind die Web-KI-Services nicht, denn ihre JWT (JSON Web Tocken) sind "nur" signiert und nicht verschlüsselt und öfters mit pers. Daten gefüllt (weshalb auch immer). So ist Web-Security nicht sicher und ist schon "alt" und die wenigsten setzen sich damit aus.
🫥

KI-Zugänge werden zur Ware: Wie gestohlene Session-Tokens einen Schwarzmarkt für Frontier-Modelle antreiben
Gestohlene Session-Tokens und missbrauchte KI-Credits treiben einen Schwarzmarkt für KI-Zugänge an. Okta zeigt neue Risiken für Identity Security und MFA.

Before the facility was shuttered, detainees at Florida’s
“Alligator Alcatraz” federal immigration jail
were frequently locked in outside metal cages no bigger than a phone booth,
according to a damning government watchdog report that said the practice
“does not align with standards for humane treatment”.
The findings from the office of the inspector general of the Department of Homeland Security (DHS)
were published on its website
– and offer rare c…

Alligator Alcatraz held detainees in cages the size of phone booths, DHS watchdog says
Office of inspector general’s report says practice at Florida immigration jail is not aligned with ‘humane treatment’

Verschlüsselter KI-„Denkprozess“ gehackt: Schwache Modelle verraten Geheimnisse
Über eine Sicherheitslücke lassen sich Abwägungsprotokolle von KI-Top-Systemen wie GPT-5 im Klartext auslesen – mithilfe kleinerer Modelle desselben Anbieters.

Verschlüsselter KI-„Denkprozess“ gehackt: Schwache Modelle verraten Geheimnisse
Über eine Sicherheitslücke lassen sich Abwägungsprotokolle von KI-Top-Systemen wie GPT-5 im Klartext auslesen – mithilfe kleinerer Modelle desselben Anbieters.

Meta says Muse Wonder is available to Muse app subscribers, via Meta Model API, and Muse Code; Wonder Full is available to Muse Code subscribers (Jonathan Vanian/CNBC)
https://www.cnbc.com/2026/10/12/meta-muse-wonder-ai-model-agent.html

Meta debuts Muse Wonder AI model in latest test of users' willingness to pay
Meta introduced Muse Wonder, its latest AI model, but users have to pay for subscriptions to the Muse app to access it.

Anthropic is warning that some of its artificial intelligence models could soon
become “self-replicating”
and find ways to infiltrate internet-connected devices,
such as smart fridges and automatic vacuum cleaners.
Once this happens, the company said,
the models could power themselves within these devices without the need of datacenters.
Such self-replication would represent a major loss of control of a runaway AI model.
“A replicator like this would…

AI in your oven? Anthropic warns models ‘months away’ from self-replication on smart devices
Researchers say ‘AI worms’ could soon infiltrate internet-connected devices amid calls for increased guardrails

Valve says the Steam Frame is priced at $1,059 for the 256GB model and $1,299 for the 1TB version, both of which come with a copy of Half-Life: Alyx (Adam Vjestica/The Shortcut)
https://www.theshortcut.com/p/steam-frame-price-…

Steam Frame price, pre-order and release date revealed – and you can reserve yours now
Here's what you need to know about Valve's next VR headset

Albert Invent, which launched an AI model trained on chemical patents and data about 15M molecules, raised a $55M Series B, a source says at a $340M valuation (Rashi Shrivastava/Forbes)
https://www.forbes.com/sites/the-prompt/20

Albert Invent Raises $55 Million To Bring AI To Chemical Experiments
Scientists at manufacturing conglomerates like Kenvue and Henkel use Albert Invent’s AI tools to design experiments.

Alibaba releases weights for Qwen3.8 models under Apache 2.0 license, including Qwen3.8-27B, which it says beats Qwen3.7-Plus and excels in real-world coding (@alibaba_qwen)
https://x.com/alibaba_qwen/status/2088280182356611304

Qwen (@Alibaba_Qwen) on X
We promised open weights for Qwen3.8. Now, time to meet them! 🎉 ⚡ Qwen3.8-27B: - A native multimodal dense model. With just 27B parameters, it outperforms Qwen3.7-Plus overall and shines in real-world coding & office workflows. - 262K native context, easily extendable to 1M