Microsoft is set to release its AI vulnerability detection tool, Project Perception, by the end of July. The tool is designed to compete with Anthropic's Mythos model, offering a more cost-effective solution for identifying vulnerabilities in AI systems. This move marks Microsoft's continued expansion into AI safety and security, aiming to provide robust tools for developers and organizations.