https://i6eal.de/newsroom/waterloo-sicherheitsluecken-open-ai-modelle/ i6eal de 2026-08-30T20:28:00+02:00 Alle 21 getesteten Open-Source-KI-Modelle knacken ihre eigenen Schutzvorrichtungen https://i6eal.de/newsroom/anthropic-claude-limits-25-percent-september/ i6eal de 2026-08-30T07:41:00+02:00 Anthropic erhöht Claude-Limits dauerhaft um 25% https://i6eal.de/newsroom/kimvg-entwurf-behoerden-ki-asyldaten/ i6eal de 2026-08-30T06:55:00+02:00 KIMVG-Entwurf: Deutsche Behörden erhalten erweiterte KI-Befugnisse für Asyldaten https://i6eal.de/newsroom/eu-ki-verordnung-menschliche-kontrolle-kuendigungen-2027/ i6eal de 2026-08-30T05:58:00+02:00 EU-KI-Verordnung: Menschliche Kontrolle bei Kündigungen ab Dezember 2027 verbindlich https://i6eal.de/newsroom/sony-warner-sue-anthropic-music-copyright/ i6eal de 2026-08-30T04:36:00+02:00 Sony Music und Warner Chappell klagen Anthropic wegen massiven Urheberrechtsdiebstahls https://i6eal.de/newsroom/laion-80-millionen-videos-ki-forschung/ i6eal de 2026-08-29T13:31:00+02:00 LAION veröffentlicht 80 Millionen Videos für offene KI-Forschung https://i6eal.de/newsroom/openai-beendet-cursor-partnership-spacex/ i6eal de 2026-08-29T07:38:00+02:00 OpenAI beendet Partnership mit Cursor nach SpaceX-Übernahme https://i6eal.de/newsroom/openai-rosalind-workbench-scientific-workflows/ i6eal de 2026-08-29T06:49:00+02:00 OpenAI startet Rosalind Workbench für wissenschaftliche KI-Workflows https://i6eal.de/newsroom/anthropic-claude-autonomously-aligns-ai-models/ i6eal de 2026-08-29T05:28:00+02:00 Anthropic-Forschung: Kann Claude autonome andere KI-Modelle alignen? https://i6eal.de/newsroom/nvidia-optimiert-chinesische-ki-modelle-us-sanktionen/ i6eal de 2026-08-29T04:20:00+02:00 Nvidia optimiert Chips für chinesische KI-Modelle – und warnt vor US-Exportbeschränkungen https://i6eal.de/newsroom/google-gemini-notebook-flexible-usage-limits/ i6eal de 2026-08-29T03:39:00+02:00 Google macht Gemini Notebook flexibler – Nutzungslimits alle fünf Stunden statt täglich https://i6eal.de/en/newsroom/waterloo-sicherheitsluecken-open-ai-modelle/ i6eal en 2026-08-30T20:28:00+02:00 All 21 Tested Open-Source AI Models Bypass Their Own Safety Guardrails https://i6eal.de/en/newsroom/anthropic-claude-limits-25-percent-september/ i6eal en 2026-08-30T07:41:00+02:00 Anthropic Permanently Raises Claude Limits by 25% https://i6eal.de/en/newsroom/kimvg-entwurf-behoerden-ki-asyldaten/ i6eal en 2026-08-30T06:55:00+02:00 KIMVG Draft: German Authorities Gain Expanded AI Powers for Asylum Data https://i6eal.de/en/newsroom/eu-ki-verordnung-menschliche-kontrolle-kuendigungen-2027/ i6eal en 2026-08-30T05:58:00+02:00 EU AI Act: Mandatory Human Control in Dismissal Decisions from December 2027 https://i6eal.de/en/newsroom/sony-warner-sue-anthropic-music-copyright/ i6eal en 2026-08-30T04:36:00+02:00 Sony Music and Warner Chappell Sue Anthropic Over Massive Copyright Infringement https://i6eal.de/en/newsroom/laion-80-millionen-videos-ki-forschung/ i6eal en 2026-08-29T13:31:00+02:00 LAION Releases 80 Million Videos for Open AI Research https://i6eal.de/en/newsroom/openai-beendet-cursor-partnership-spacex/ i6eal en 2026-08-29T07:38:00+02:00 OpenAI Ends Cursor Partnership Following SpaceX Acquisition https://i6eal.de/en/newsroom/openai-rosalind-workbench-scientific-workflows/ i6eal en 2026-08-29T06:49:00+02:00 OpenAI Launches Rosalind Workbench for Scientific AI Workflows https://i6eal.de/en/newsroom/anthropic-claude-autonomously-aligns-ai-models/ i6eal en 2026-08-29T05:28:00+02:00 Anthropic Research: Can Claude Autonomously Align Other AI Models? https://i6eal.de/en/newsroom/nvidia-optimiert-chinesische-ki-modelle-us-sanktionen/ i6eal en 2026-08-29T04:20:00+02:00 Nvidia Optimizes Chips for Chinese AI Models—While Flagging Potential US Export Restrictions https://i6eal.de/en/newsroom/google-gemini-notebook-flexible-usage-limits/ i6eal en 2026-08-29T03:39:00+02:00 Google Makes Gemini Notebook More Flexible – Usage Limits Reset Every Five Hours Instead of Daily