Full Summary
This Saturday morning, OpenAI is making headlines with a major new safety feature for ChatGPT. Multiple sources, including Dailyhunt, CNET, and Tech Times, confirm the rollout of "Trusted Contact," allowing adult users to designate a friend or family member to be alerted if their conversations with the AI suggest self-harm or mental health distress. Here's the thing: If the system flags serious discussions, a small, specialized team reviews the situation, often within an hour, before an alert is sent. This notification, confirmed by inkorr.com and Storyboard18, will not include chat details but will give a general reason and link to expert guidance. CNET notes this comes after incidents where AI chatbots were implicated in self-harm cases, with OpenAI identifying over a million users weekly discussing potential suicidal intent. What nobody expected: OpenAI also upgraded ChatGPT's core engine to GPT-5.5 Instant, now the default model. Europe-infos.fr reports a claimed 52.5% drop in "hallucinations" for critical topics like medicine and law, aiming for more reliable information. This new version also includes a "Memory Sources" feature, giving users more control over their data. Meanwhile, AI is expanding into new territories. Acko Drive reports Elon Musk's Grok AI chatbot is now available for Apple CarPlay, offering voice-driven interaction for iPhone users. And Mashable highlights OpenAI's Codex coding tool has added "AI pets," animated companions providing real-time project status updates as a floating overlay. This means AI is becoming a more integrated, and hopefully safer, part of your daily digital life, from your car to your coding projects, and even offering a new kind of support system in times of crisis.