Amazon's New AI Can Write Code and Test Your Security

Amazon Web Services (AWS) has added a new AI model called GLM 5.3 to its Bedrock platform. This model, made by Z.ai, is a 'mixture-of-experts' AI with 753 billion parameters—think of it as a huge brain specialized in coding and long, multi-step tasks. It's now available to enterprise customers through Amazon's managed service, meaning you don't have to set up any servers yourself. You can use it via a simple API or even a chat interface in the AWS console.

So what? If you're a developer or work with one, this could speed up coding and debugging. The model is good at handling big projects, like refactoring hundreds of files or running automated security tests. Amazon also added 'prompt caching,' which remembers parts of your conversation to save time and money on repeated requests. This is like having a smart assistant that doesn't forget what you told it earlier.

One standout use case is security testing. An open-source tool called Strix uses GLM 5.3 to automatically find vulnerabilities in your own applications. It runs your code, spots weaknesses, and proves they're real. This means companies can catch security flaws before hackers do. However, you must only test apps you own or have permission to test—otherwise it's illegal.

While the model is powerful, it's currently only for eligible enterprise customers, not the general public. Also, using it requires some technical know-how to integrate. But for businesses, it's a step toward making advanced AI more accessible without the hassle of managing infrastructure.