Microsoft AI Model: Finds Cyber Flaws, Lowers Cost

Microsoft has introduced its first artificial intelligence model specifically designed to find security flaws in source code. This new model is called MAI-Cyber-1-Flash. It will work with OpenAI's GPT-5.4 within Project Perception, a system of AI agents built to detect and repair vulnerabilities. The combined system outperformed other security models on the CyberGym benchmark. It also operated at about half the cost. Project Perception will enter public preview on August 3rd. These tools can recommend code changes and apply fixes with customer approval. They can also connect with products outside Microsoft's ecosystem. This launch is Microsoft's first major cybersecurity initiative since February. The company aims to improve efficiency as cyberattacks become faster and more sophisticated. This could mean more secure software for everyone.