Powered by GPT-4: Microsoft’s new AI Security Copilot to enhance cybersecurity

According to Microsoft, 1,287 password attacks occur every second around the world.
Mrigakshi Dixit
Microsoft Romania headquarters.
Microsoft Romania headquarters.


Microsoft is now focusing on cybersecurity as part of its ongoing efforts to incorporate generative artificial intelligence into the majority of its products. The company previously announced an AI-powered assistant for Office apps. 

According to Microsoft, 1,287 password attacks occur every second around the world. And the number of attacks has increased by 67 percent in the last five years.

To enhance cyber security, Microsoft Corp has announced the implementation of the next generation of AI in its security products. Microsoft Security Copilot would help professionals detect and respond to threats more quickly.

“Today the odds remain stacked against cybersecurity professionals. Too often, they fight an asymmetric battle against relentless and sophisticated attackers. With Security Copilot, we are shifting the balance of power into our favor. Security Copilot is the first and only generative AI security product enabling defenders to move at the speed and scale of AI,” said Vasu Jakkal, corporate vice president of Microsoft Security, in a statement

The new assistive tool is powered by OpenAI's GPT-4 generative AI, and Microsoft's security-specific model based on daily activity data. This security model collects approximately 65 trillion threat signals per day, which could serve as valuable intelligence data for this tool.

Reportedly, it will function as a simple AI assistant, much like a chatbox.

Security Copilot is designed to amplify the capabilities of security teams. This tool can help security professionals with demanding security tasks and applications such as investigating security incidents, quickly summarizing events, and creating a quick report. 

According to the official release, it is intended to help security teams analyze what is going on in their environment, learn from existing intelligence, correlate threat activity, and make more informed, efficient decisions at machine speed.