Each week, we bring you 5 stories that resonated the most in our internal Slack channel #AI-news. We write the newsletter using various AI tools because we're an AI company and our marketing wants to move with the times too. 😎
Today you're reading the 68th issue in a row.
#1
OpenAI and FDA 🤝: In talks about using AI in drug approval 💊🤖.
OpenAI has reportedly met 🤝 with the U.S. Food and Drug Administration (FDA) 🇺🇸💊 to explore how artificial intelligence 🤖 could assist in evaluating new medications. The discussions center around a tool with the working name cderGPT 🧠, named after the Center for Drug Evaluation and Research—the FDA division responsible for most prescription and over-the-counter drugs 🧾.
The drug approval process in the U.S. is notoriously slow 🕒—often taking more than 10 years. AI could help especially in the later stages 📂📋, where the process involves the most paperwork and regulatory checks 🔍. While these are still early-stage talks 🗂️, the goal is to get safe and effective medications to patients faster 🏥⚡ than is currently possible.
The FDA, however, remains cautious ⚠️. AI must not make faulty or unreliable decisions ❌—especially when human health is on the line ❤️. That’s why safety and trustworthiness 🔒 will be absolutely critical 🎯 in how any AI tool is integrated into the drug review process.
#2
LegoGPT 🧱: AI designs buildable and stable LEGO models 🤖🎨.
Researchers at Carnegie Mellon University 🎓 have developed LegoGPT—an AI-powered tool 🤖 that can generate real, buildable LEGO models 🧱 based on simple text prompts. Just type something like “LEGO guitar” 🎸 and the tool will automatically create a detailed building guide. No manual needed—you get everything straight from the AI 🧠✅.
The model was trained on tens of thousands of real LEGO constructions 🏗️ and includes validation mechanisms ⚙️ to ensure that the final design is physically buildable (no floating bricks 🪄). Even better—LegoGPT is open-source and free to use 🆓. With a few tweaks, you can even connect it to a camera 📷 that scans your current brick collection and suggests models you can build with what you already have 🧱✨.
It’s a great example of how AI can blend creativity with the physical world 🎨. And the best part? You can try it at home 🏡—all you need is LEGO and a bit of creative spirit 💡👾.
#3
Chip Security Act 🏛️: Protecting against chip smuggling to China 🇺🇸🇨🇳.
Senator Tom Cotton 🏛️ has introduced a bill called the “Chip Security Act,” which would require high-performance AI chips 💽 subject to export restrictions 🚫 to include built-in location tracking 📍. The goal is to prevent these chips from being illegally shipped to countries like China 🇨🇳, where they could potentially be used for military purposes. Companies would also be required to report if their chips show up where they shouldn't—or if the hardware has been tampered with ⚠️.
The bill comes in response to growing concerns 😟 following reports that Nvidia chips are being smuggled into China despite the restrictions 📦🔍. It’s part of a broader U.S. effort 🇺🇸 to control where American technology ends up and how it’s used ⚙️—especially in sensitive sectors like AI and defense 🛡️.
Interestingly, a similar bill is being prepared by a Democratic lawmaker 🗳️. This suggests that AI hardware security 🔐 has bipartisan support and may soon become a key focus of U.S. tech policy 📈.
#4
OpenAI and Microsoft 🤝: In strategic talks about the future of their partnership 💼📊.
OpenAI and Microsoft 🤝 are currently engaged in tense negotiations to redefine their partnership 📝. Microsoft has already invested $13 billion 💰 into OpenAI, but now that OpenAI is transitioning into a for-profit Public Benefit Corporation (PBC) 🏢💚, both sides are reassessing how the collaboration will continue—especially around ownership, control, and access to technology 🔧📊.
Microsoft is reportedly seeking a larger stake 📈 or at least an extension of its rights to use OpenAI’s technology beyond 2030 📅, when their current agreement ends. Meanwhile, OpenAI is gradually becoming more independent 🔓 and, in some areas, even emerging as a competitor 🧠⚡—such as in enterprise services or through its own AI infrastructure project, Stargate 🌐.
These talks go beyond contracts and numbers 💼📄. What’s at stake is how much influence Microsoft will retain over the future of one of the world’s most influential AI companies 🌍—and whether it can stay in the game 🕹️ as the rules continue to evolve 🔁.
#5
Apple and Safari 🍏: Considering integration of AI search engines like Perplexity 🤖🔍.
According to senior executive Eddy Cue 🧑💼, Apple 🍏 is considering adding new AI-powered search engines 🤖🔍—such as Perplexity, OpenAI, or Anthropic—to Safari. Why? For the first time in 22 years, the number of traditional searches through Safari has declined 📉. That’s a problem for Apple, since Google reportedly pays around $20 billion per year 💰 to be the default search engine in Safari. Fewer searches mean less revenue 💸.
While Cue says today’s AI search engines aren’t quite ready to replace Google yet ⚖️, Apple is keeping its options open 🚪. If a true competitor emerges, Apple wants to be ready to shift directions 🔄.
ChatGPT is already integrated into Siri 🎙️, and Google is working to bring Gemini to iPhones 📱. So it’s clear Apple doesn’t want to be left behind 🚫 as AI becomes the new gateway to information 🌐🧠. Money, technology, and the future of web search are all on the line 🔍💼.
