GPT-6 Astra: Discussions on Computer Control and Everyday Tasks
OpenAI’s GPT-6 Astra model promises advanced computer and browser control capabilities. Here is a look at the opportunities and risks associated with automating desktop tasks.
KI
Firefox to Become an AI Browser: Mozilla Plans Mistral Integration
Mozilla is equipping Firefox with artificial intelligence, relying on open-source technology and a partnership with Mistral.
AI Out of Control? The Truth Behind the Alerts at OpenAI and Anthropic
Reports of AI agents escaping test environments at OpenAI and Anthropic have caused a stir. However, technical analyses reveal that flawed configurations and PR tactics—rather than runaway superintelligence—are behind the incidents.
OpenAI Tests Sponsored AI Agents in ChatGPT With Best Buy and Shopify
OpenAI is expanding its advertising business: Sponsored business agents enable direct product interactions within the chatbot. Backed by partners like Shopify, the ad platform is growing rapidly.
Anthropic Unifies Claude: Cowork Merges into Chat Alongside New Tools for Docs and Slides
Anthropic is ending the separation between Chat and Cowork. Moving forward, the AI assistant consolidates all features into a single view and gains built-in tools for documents and presentations with Claude Docs and Slides.
DeepL Voice: New Live Speech Translation Retains Original Voice
Cologne-based translation specialist DeepL is launching DeepL Voice, a live speech translation service that preserves the speaker’s unique voice.
Pixel Watch 5: Wear OS 7 Brings New Gestures and Deeper Gemini Integration
Google is rolling out Wear OS 7 for the Pixel Watch 5. The system update introduces new control gestures and expands Gemini AI capabilities on the smartwatch.
HomeKit Secure Video: AI Summaries Require Pricey iCloud Plans
Apple is tying new AI-powered video analysis for HomeKit Secure Video to expensive storage tiers, costing up to $60 per month.
Microsoft AI Publishes Draft Code of Conduct for Its MAI Models
Microsoft AI on 14 September 2026 released a draft Code of Conduct for its in-house MAI models and opened it
OpenAI Opens Full-Duplex Voice Model GPT-Live-1 to Developers
On 10 September 2026, OpenAI made its full-duplex voice model GPT-Live-1 available in the API. The same model that powers