ChatGPT Health Now Available to All US Users

On July 23, OpenAI began offering its health information feature, ChatGPT Health, to all users in the United States. The feature lets users share their health data with ChatGPT and ask health-related questions or receive advice. Initially available to a limited audience, the nationwide rollout now makes it accessible to a much larger user base.

As AI's role in healthcare continues to expand, ChatGPT could serve as an entry point for health consultations — but concerns about inaccurate medical advice remain strong. What safety measures OpenAI has put in place will be a key focus going forward.

Google's Gemini Nears 1 Billion Users

It has been revealed that Google's AI assistant, Gemini, is approaching 1 billion users. Google has already integrated Gemini across multiple services including Search, YouTube, and Android, making it increasingly likely that Gemini will join the ranks of the company's vaunted "billion-user products."

Gemini's rapid growth suggests that AI assistants are becoming firmly established as everyday tools. That said, more detailed disclosure is needed on how users are counted — active users versus registered accounts.

Vulnerability Lets a Tampered ChatGPT Link Run a Rogue AI Agent

A vulnerability has been reported in which simply opening a single tampered ChatGPT link allows a rogue AI agent — one that receives instructions from an attacker every five minutes — to infiltrate a corporate network. Reported by both The Decoder and The Register, the issue highlights the risk of ChatGPT's custom GPTs and agent features being abused.

By manipulating a link's parameters and embedding malicious prompts, attackers can launch the rogue agent without the user ever noticing. For enterprise security, validating the URLs and parameters of AI tools has become a new challenge.

Microsoft Replaces OpenAI Image Models in PowerPoint and Bing

According to Bloomberg, Microsoft has begun replacing the OpenAI image generation models it previously used in products such as PowerPoint and Bing with alternative models. The move is drawing attention as a possible sign of change in the partnership between the two companies.

While Microsoft is one of OpenAI's largest investors, it has also been strengthening its own in-house AI, and is believed to be rethinking its model choices from the perspectives of cost reduction and performance optimization. How the relationship between the two companies evolves going forward is a topic of considerable interest.

Nvidia Plans to Send GPUs to the Moon

Nvidia is developing and deploying GPUs for space, and it has now been revealed that the company plans to send GPUs to the lunar surface. Reported by TechCrunch, the news shows that demand for AI computing power is expanding beyond Earth into the field of space exploration.

AI processing on the Moon would solve communication latency problems and enhance the autonomy of probes and rovers. The fusion of space development and AI is expected to open up a new frontier.

Meta Unveils AI Optimism Ad Set to a Song About Human Extinction

Meta has launched a new advertising campaign delivering an upbeat message about the future of AI. However, the ad has drawn attention for an unintended irony: it uses a song about the extinction of humanity. According to TechCrunch, the ad was intended to ease anxiety about AI, but the choice of music ended up inviting criticism.

It's an episode that illustrates how hard AI companies are working to shape their public image.

AMD Announces New Products Targeting a $2 Trillion Market

Bloomberg reports that AMD has announced a new lineup of products aimed at strengthening its competitive position in the AI chip market. The company believes the market for GPUs and accelerators used in AI inference and training could grow to as much as $2 trillion in the future.

In an AI chip market where Nvidia holds an overwhelmingly dominant share, how much of a shift AMD's pursuit will bring about remains unknown — but more options are a win for users.

---