Microsoft Project Perception: Microsoft is coming up with a new AI agenting security system, which will be able to detect cyber attacks before they happen and provide protection.
Microsoft Project Perception: With the advent of AI, cybersecurity is facing new challenges. Hackers are using AI to launch attacks that are now increasingly difficult to prevent. Individual users and even large companies have been unable to escape these challenges. Human methods are proving insufficient to counter these attacks, and now AI will be the only shield against them. Microsoft is now launching Project Perception. Let's explore this project and how it can protect against AI-induced cyberattacks.
What is Microsoft's Project Perception?
Project Perception is an AI-based security system that reacts at the same speed as the threat. This system won't just send alerts to security teams. As soon as a threat is detected, it will begin to think independently and, using AI, will continuously monitor the threat, taking action if necessary. Instead of operating on a single system, it will operate across the company's entire digital network.
How will Project Perception work?
Three types of AI agents will operate within this system . Red team agents will look for flaws in the system that could be exploited by attackers. Next, the blue team will come into play. These agents will identify the flaws and determine what is causing the concern. Green team agents will then begin their work and fix those flaws. These three teams will work in a loop, learning incrementally over time. This will ensure complete security for the system, and all this work will be done under human supervision, with humans having the authority to make the final decisions.
Why does Microsoft trust this system?
Microsoft says this system can monitor devices, apps, data, and cloud systems. It can then take action based on what it detects. When combined with a multi-model setup, it can select different AI models for each task. This will ensure that a single AI model doesn't have to handle all the work. Microsoft has built this system on its new Cyber Stack. This is a chain that converts signals into context. This context is then passed to models and agents, which take action. Together, these elements can provide a reliable security layer.
Cyber attackers are using AI to perform everything from identifying system vulnerabilities to executing full attacks. The first step is to identify system vulnerabilities, which has become easier and faster thanks to the advent of AI. Next, information about the target is gathered from social media platforms and other public sources. This information is analyzed using AI to create personalized and relevant messages, which then initiate phishing and other attacks.
How many types of AI-powered cyber attacks are there?
AI-driven phishing and social engineering attacks - These involve manipulating human behavior to gather information. Voice cloning and other similar attacks are used in such attacks.
Deepfakes - Fake videos created using AI to simulate real-life images are being used for fraud.
Adversarial AI - In this, attackers can disrupt the performance and accuracy of AI through misinformation or manipulation.
Ransomware - AI ransomware is a type of ransomware that continuously improves its performance with the help of AI and automatically devises ways to evade defences. Therefore, it becomes very difficult to avoid it.
