Full Summary
This Thursday morning, widespread outages have hit major AI tools, with OpenAI confirming elevated errors affecting both ChatGPT and Codex. PCMag UK and the International Business Times, Singapore Edition, both report that other leading AI services, including Grok, Claude, Google’s Gemini, and Microsoft’s Copilot, also experienced issues, with SpaceXAI blaming a problem at its Memphis compute center for the Grok outage. While OpenAI's status page now indicates systems are fully operational, some users may still see isolated problems. In other news, OpenAI is launching GPT-6 Astra, a major upgrade to ChatGPT and Codex, which 9to5Mac describes as its "most intelligent and aligned model." Astra scores 98% on FrontierMath Tier 4 and 99.9% on ARC-AGI-3, and achieves a 100% score on ExploitBench. This new model improves computer use, letting ChatGPT utilize your Mac or other desktop, even in the background. Android Authority also confirms that Claude Cowork offers similar background computer use, currently in beta for Pro and Max plans on the Claude app for Mac. On the consumer front, both Geeky Gadgets and geeky-gadgets.com confirm that with the release of iOS 27, users can now replace Siri with ChatGPT for text-based queries, offering more detailed explanations and creative suggestions, though Siri remains better for device-specific tasks. Sonos is also integrating AI and ChatGPT into its new Sonos 27 operating system, allowing external AI assistants to control Sonos systems, according to gizbot.com. For younger users, OpenAI has launched a version of ChatGPT specifically for teens. The Boar reports this new model includes stricter age verification, parental controls, and a "study mode" for step-by-step problem-solving. It also has stronger safety settings to flag conversations about self-harm or romantic roleplay, aiming to prevent teens from forming human attachments to the AI. Finally, the U.S. Department of Defense has added OpenAI's "ChatGPT Mil" and "Grok for Government" to its GenAI.mil platform, ending its exclusive reliance on Google Gemini. XenoSpectrum notes this creates a multi-model system, allowing military and civilian personnel to choose between different AI models. These developments mean that from your next iPhone update to how your smart speakers function, AI is becoming more integrated into daily life, offering both enhanced capabilities and new security considerations.