What is our primary use case?
Microsoft Security Copilot capabilities are being attached to managed services. In every single one of the services including
Microsoft Purview,
Microsoft Sentinel, and in the
XDR stack with E5 entitlements,
Microsoft Security Copilot is being attached to each one of these managed offerings to add reduced time to detect, respond, and derive additional insight that most companies cannot purchase a full-time employee for. A full-time employee is usually too expensive, and Microsoft Security Copilot, especially with recent changes, allows for it to be a more beneficial route for some companies, specifically government.
How has it helped my organization?
The impact of the integration of Microsoft Security Copilot with other Microsoft security solutions, such as
Purview, on the organization's security posture is that it has had a rebounding effect. Because everybody wants to use it, what it has done is forced the entirety of organizations leveraging Copilot in general to actually adopt
Purview. In order to use it, they have realized their security and data governance is not where it needs to be. This has magnified the already understood security concerns that most legal and compliance teams already had and made them now a security problem, which if you have worked in the IT field, you know if security has a problem, they are most likely going to tackle it with some level of vigorous attempt.
Microsoft Security Copilot has helped reduce the mean time to resolution. The mean time to detect is still dependent on the solution itself, but there needs to be a new acronym for MTTU, the mean time to understand and determine what is actually happening. Resolution is an outcome of that, obviously, but there is a layer in between that even a junior analyst can now respond to an alert.
What is most valuable?
The ability to build Microsoft Security Copilot agents is outstanding. The first-party agents are amazing, and the ability to derive additional insights and create reduced queues of information and alerts, stating which ones need focus, is valuable. The additional feedback that can be given to the agent, which allows for it to refine its process, is outstanding in the current state. Now that the cost has matched the value, AI in Microsoft security cannot be avoided anymore. It is here now.
What needs improvement?
The natural language interface is extremely approachable because inputs are as good as outputs, which is a universal understanding of natural language models and LLMs with that interface. The real value is in the tools that sit underneath it. The question is what the agent can actually do once asked that question. Can it pull information from areas that previously it could not? Can it actually action on these? Can one determine, via MCP or something of that nature, what the security posture should become and what steps need to be taken next? It is truly dependent on the underlying infrastructure and technologies, less about the interface itself.
AI features have not quite been used to automate tasks yet. Having a human in the middle is still proving to be the best method for adoption because not many people are ready to give AI the wheel. Human involvement with enough refinement and inputs from that trusted human, and then eventually, companies should allow AI to fully automate their security processes within about a year.
Agent-to-agent is how Microsoft Security Copilot can be improved. Agent-to-agent is the next logical step in creating an automation chain where RBAC can be enabled to ensure that each agent only has a specific amount of permissions through least privilege and zero trust principles. The capability to then understand, detect, determine, and respond can be reduced even further.
For how long have I used the solution?
There has been involvement in Microsoft Security for about eight years and in security as a whole for about ten years. When starting out,
Microsoft Defender for Endpoint was not that great because it was brand new around the 2016 era.
Sentinel had just released, and Defender for Endpoint was not really the top of the market best of breed, but the vision was clear that Microsoft was going to get it right eventually. The dedication to this journey has continued.
What do I think about the stability of the solution?
There has not been enough stability and reliability demonstrated yet. The hosting method is not known, which is a black box situation. When talking about confidentiality, integrity, and availability, availability is the critical concern. When relying on Microsoft Security Copilot as an organization and it goes down, the question becomes what happens then. How do you know it is going to fail over appropriately? Is it hosted in one area? Is it hosted in multiple areas? Is it distributed? None of that is known yet, and it does not appear that this has been revealed to the masses. However, as custom agents are built, organizations will be able to control that, which makes agent-to-agent even more important.
What do I think about the scalability of the solution?
Microsoft Security Copilot is scalable for a growing organization. Scalability is the name of the game, involving understanding exactly where focus and money need to be placed and ensuring that where focus and money are placed can scale with the size, speed, and capabilities of organizations as they adopt AI in the workplace.
How was the initial setup?
The experience of deploying it is that if you know what you are doing and know where to go, it is very easy and literally a click of a button. If you do not, then there is apprehension. Education is going to be extremely important in the near future.
There is not enough documentation, and visual learning resources are needed. A 15-second video showing the process of pressing the button to get started would be helpful. The plan is to personally create those resources to show organizations exactly how to press that button without fear and what the repercussions and costs to the organization might be, ensuring protection so that a boss, CISO, or CFO does not come asking why there is a ten-thousand dollar bill.
What's my experience with pricing, setup cost, and licensing?
Regarding the pricing, setup cost, and licensing, it was unreasonable before and was not realistic. No one was going to use it, and no one really did use it unless given some money to blow in the first month just trying to understand how to use it. The announcement this week has changed everything. Microsoft Security Copilot is now integrated with E5, providing four hundred SKUs for one thousand users who have E5, up to ten thousand users. If buying E5, everything is now included.
What other advice do I have?
AI in Microsoft Security Copilot for incident response is effective. Full-time employees represent a longstanding known problem that most organizations are not serviced enough with or do not have enough to actually handle the volume that they receive. The goal of Microsoft Security Copilot is to bridge that gap and solve a problem that has been known to exist for a long time. The review rating for this product is 7.