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 77th issue in a row.
#1
OpenAI launches GPT-5 for everyone 🤖 – faster, smarter, more reliable 🚀.
OpenAI has officially released GPT-5 to all ChatGPT users and developers via API 🚀. The new model is faster, more accurate, and significantly better at logical reasoning, programming, and solving multi-step tasks 🧠. There’s no longer a need to switch between modes—it’s all unified into one system 🔄.
GPT-5 currently leads benchmarks in coding, writing, and medical advice, and it can generate functional applications within seconds 🖥️. OpenAI has also improved its safety mechanisms—new “safe completions” ensure helpful responses without the risk of harmful content 🛡️.
The model marks an important step toward artificial general intelligence, though GPT-5 still can’t learn from new data after release, and the company has not disclosed its training data 🔍.
#2
Apple AI to get GPT-5 upgrade 🧠 with new systems this fall 🍂.
Apple has confirmed that the ChatGPT integration in Apple Intelligence will switch to the new GPT-5 model, but only in the fall with the release of iOS 26, iPadOS 26, and macOS Tahoe 26 🍂. These updates will introduce improved AI features, such as a smarter Siri and more advanced image analysis 🧠🖼️.
Until then, Apple devices will continue using the GPT-4o model 📱. It's still unclear whether GPT-5 will appear in beta versions of the systems — Apple hasn’t provided any information about that, which is creating uncertainty among developers and testers 🤷♂️.
However, the switch to GPT-5 could significantly improve the accuracy and capabilities of Apple AI, especially in areas like logical reasoning and programming 💡💻.
#3
Google is investing $1 billion into AI education at U.S. universities and nonprofit organizations 💰.
Students will get free access to Gemini tools, and researchers will receive cloud credits ☁️🔬. Over 100 schools have already joined the program, with the goal of eventually covering all accredited nonprofit colleges in the U.S. 🏫🇺🇸.
The company says it wants to support innovative approaches to AI education in collaboration with academia 🤝. However, concerns are emerging—including the risk of cheating, declining teaching quality, and the growing influence of Big Tech firms on education ⚠️🏢.
Other players like Microsoft and OpenAI are launching similar initiatives 📊, marking a new phase in the competition to attract future AI talent 🚀👩🎓.
#4
CEO of DeepMind 🤖: AGI could emerge within 10 years 🔮.
DeepMind CEO Demis Hassabis claims that artificial general intelligence (AGI) could emerge within 5 to 10 years and trigger transformative changes—according to him, up to ten times greater and faster than those of the Industrial Revolution 🤖. He believes AGI will bring extreme productivity and “radical abundance,” but the key issue will be the fair distribution of those benefits 🧮.
Not everyone shares his optimism ⚠️. Critics point out that Hassabis, as the head of one of the most influential AI firms, has incentives to exaggerate progress—and that AGI development still faces serious technical limitations 🧱. Concerns also remain around control, jobs, and societal inequality 📉.
Nonetheless, with these statements, Hassabis joins a growing number of AI leaders who predict a rapid arrival of general artificial intelligence—even though the consequences of such a shift remain deeply uncertain 🔮.
#5
Nvidia and AMD can sell chips to China again 💰 – for a 15% fee 🇨🇳.
According to the Financial Times, Nvidia and AMD will once again be allowed to sell their most powerful AI chips to China—in exchange for handing over 15% of the revenue from those sales to the U.S. government 💰. This applies to new export licenses for chips like the Nvidia H20 and AMD MI308, which were originally banned due to national security concerns 🛑.
This reversal follows a temporary U.S. ban on exporting cutting-edge AI technology to China 🌐. Nvidia pledged billions in domestic investments, which officials say helped ease the restrictions 🇺🇸. It’s also part of broader trade negotiations, including deals involving rare metals critical to semiconductor production ⚙️.
Some analysts warn that while the agreement boosts oversight of exports, it could also undermine the original security goals ⚖️. AI chips remain a strategic technology, and exporting them to China continues to be a controversial issue 🔍.
