This weekly update highlights major scientific breakthroughs in agentic AI, which is artificial intelligence that can make decisions and complete tasks on its own. The most exciting news comes from the mobile technology conference in Barcelona, Spain, where Huawei showed off its new Agentic Core system that helps AI agents respond almost instantly by cutting wait times by 90%. Another important announcement happened in the United Kingdom, where the government said it would spend £40 million (which is about $50 million) on a new research lab focused on Structured Language Models—special types of AI that can think through problems step by step.

In addition to hardware improvements, companies are rethinking how they build software. Instead of adding AI to old systems, they are now designing new systems from the ground up using AI agents. This agentic-first approach means that AI agents handle the hard work while humans focus on bigger decisions. Meanwhile, MIT scientists created new AI technology that can help develop medicine for cancer treatment much faster and cheaper than before.

The biggest change this week shows that scientists and businesses are moving away from just making bigger AI models and instead focusing on making smarter AI agents that can actually accomplish real tasks. These agents are being used in research, healthcare, and business, and they are changing how professionals work. However, scientists also warned that these powerful tools come with challenges, especially around security and fairness, since not everyone has equal access to these expensive technologies.