HomeTechnologyMicrosoft presents the Ire project, an AF agent capable of detecting malware

Microsoft presents the Ire project, an AF agent capable of detecting malware

Even in Prototype Stadium, this agent managed to generate a conviction “solid enough to justify an automatic block” for a malware sample.

Microsoft, Google, Openai … After the chatbots, AI companies are increasingly betting on agents. Able to perform tasks in the name of users, these systems can be useful in many areas, including cybersecurity.

On August 5, Microsoft announced Project Ire, an agent who can detect malware (“malware”). As a result of a collaboration of several teams within the company, it can analyze and classify software without human assistance.

More specifically, automates what is called retro engineering, which is “the absolute reference in terms of malware classification,” Microsoft says in a blog article. This consists in the analysis of a software to deduce its operation.

Malware and security defects

Project Ire uses tools to carry out this retro engineering and, therefore, deduce whether the software is malicious or not. Even in Prototype Stadium, he is the first Microsoft agent who has generated a conviction. “A solid enough detection to justify an automatic blockage” for a sample of advanced persistent malware (APT). An attack designed to be stealthy and continuous against a specific entity. Since then, this sample has been identified and blocked by Microsoft Defender, the free antivirus integrated in Windows.

Microsoft is not the only company that works in agents dedicated to cybersecurity. In 2024, Google announced Big Sleep, who was fed by Gemini and who can search and detect unknown security failures in the software. After detecting his first vulnerability last November, he discovered many others since then, according to the company.

On August 4, the vice president in charge of security within Google also revealed that Big Sleep had identified 20 security failures in several popular open source software.

In addition to agents, chatbots also have cybersecurity skills. Participating in the piracy of student competitions, Claude, the chatbot developed by Anthrope, has been following victories for some time. As the new company in Axios revealed, it exceeds almost all its human competitors with minimal human assistance and little or no effort.

Author: Kesso diallo
Source: BFM TV

Stay Connected
16,985FansLike
2,458FollowersFollow
61,453SubscribersSubscribe
Must Read
Related News

LEAVE A REPLY

Please enter your comment!
Please enter your name here