Microsoft Unveils Agentic Security Capabilities

Key Takeaways

Microsoft is unveiling new agentic security capabilities to its Sentinel platform hoping to offer a unified security offering in the age of AI. Joining us now to discuss this is Microsoft security corporate vice president Vasu Jakal. And there's a raft of updates Vasu to offer a more centralized platform with improved security po P co-pilot with security store offerings. How is this actually going to make it easier for your clients to identify and then act on threats? Yeah, it's great to be here with you, Caroline. We live in unprecedented times as AI becomes mainstream and we've seen the rise of this frontier firm where agents and humans collaborate and work together. But all of this is happening against a backdrop of an escalated threat landscape. We're seeing 7,000 password attacks per second. Cyber crime is costing us $9.2 trillion. There's data risk and increasing concerns about AI accessing confidential data and not to mention the complexity that defenders everywhere are facing. That's what today's announcement is about. It's about unifying security. It's about simplifying security and it's about making sure that we can defend at agentic speed. The first part of that is Sentinel. We're expanding Sentinel to be the backbone of modern agentic defense. We're introducing new capabilities like graphs which takes the trillions of signals that sentinel processes and make sure that we can convert this noise into insights and defenders can see attack and attack patterns to proactively support. And then the second part is we are expanding security copilot where we now have an agent builder that anyone can build their own agents using natural language and publish them in the security store so that it can be easily accessed. >> That backbone fasu does it not make it a single point of failure though? How are you going to be able to push back risks when you have got this unified approach in some way? We we've always believed security is a team sport and one of the advantages of having an end-to-end platform is the ecosystem that you can integrate with it. Today on average organizations have 40 plus tools and they need that end to-end visibility. So what Sentinel does is with it 350 plus connectors, it connects and brings security tools that organizations have into this data and then we're building security into it grounds up so that we are not only adhering to security practices but responsible AI practices as well and helping organizations really with that comprehensive protection. We see AI not only as the aggressor but as the defender in many ways when it comes to cyber security with everyone being able to develop their own agents. How does one keep track of that in reality? What are you going to be having to do underneath the hood? >> Well, we're very excited about the superpowers that AI brings for defenders because you're right, we need to use AI for defense and it is going to change this asymmetry in the threat landscape. Th this is why we're introducing the security store because now anyone can build an agent through security copilot and they can publish it in the security store. It enables it enables organizations to quickly discover what agents are available to them right in the tools like defender and purview that Microsoft supports. I I'm going to ask a sensitive question here, Vasu, but Microsoft has had its fair share of, shall I say, PR incidents where vulnerabilities have attacked Microsoft's own offerings and there has been concern about the level of security that has been on offer by the company. How have you managed to tackle that from a reality standpoint and from ultimately what you're seeing in the data? >> Yeah, Caroline, we have had many learnings and we've also learned that we need to defend differently in this age of AI. In November 2023, we introduced our secure future initiative to address that very thing. It is the largest cyber security initiative in the industry with the equivalent of 34,000 engineers working on it. It has a comprehensive way of looking at security on securing identities on securing production systems, engineering systems, but it's also about making sure we have 24 by7 accelerated threat response and detections. more so this has been a cultural transformation at Microsoft where we've made security our number one priority and we've made sure that security comes above all else and not only security teams but any team make sure that they have security as their top top priority. So this is some of the things that we are doing and we are seeing our customers really learn from that and also use our best practices. >> It's interesting that you talk about that culture shift. A lot of that comes down to your own talent. How much have you had to beef up talent within you know the hiring practices at the moment because all we talk of is talent wars at the moment. Is it really fierce within the cyber security area as well? >> Yeah it is cyber security is a cultural transformation as and as part of secure future initiative we've taken a big uh initiative on skilling and upskilling. We have a security academy at at Microsoft where we have we are training every single employee to learn about how to do security in this new age of AI and we also extend that training to our partner ecosystem and our customer ecosystem so we all can learn AI and we can learn security together. >> Let's talk about those partners and let's talk about the third party tools in particular. How have you gone out there and decided to work alongside you've got Zcala one trust you've got IBM as well. What sort of integrations are you seeing? What sort of best practices have you worked on? >> Yeah, we're very very grateful to our partner ecosystem. Security takes a village. It's a team sport. We have 350 plus connectors into Sentinel which is our platform for AI security. And we continue to work with Zcaler, IBM, Cisco, and many others to make sure we have the right integrations into this platform. We have 15,000 partners globally that work together with Microsoft security so that we can secure customers where they are