I know this is a security focused blog however, the future of security is using AI … and the future of AI is using agents. There aren’t any details being released in this Bloomberg post found HERE, however it’s important to see this is coming.
One current challenge with AI is knowing what is going on at the user level aka baselining and acting based on user behavior. Most AI technology is prompt based meaning you have to know what you want to ask or prebuild automation to take action. With agents, this concept becomes fuzzy. Agents could potentially monitor what is happening at the host level and like with text predictions … predict what users should be doing or need and offer input or action. It reminds me of the classic “Clippy” concept where you start writing a resume and Clippy pops up saying “looks like you are writing a resume … I have some advice that could help”. Now take this concept and ramp it up 1000 times based on the power of AI and current compute and you have the potential of AI to really become impactful in our daily lives.
Imagine a security professional working like Iron Man …. where Jarvis is seeing things along with Iron man and helping him in near real time make decisions. Jarvis doesn’t just stand by waiting to be prompted …. Jarvis jumps in based on various environmental events and helps with tasks that are better suited for machines such as large computations that Iron man needs done within seconds. For example “Jarvis …. calculate the angle I need to go to make this next turn”. In the near future, a security professional will have a form of Jarvis that jumps in and guides decisions based on available threat intelligence, alerts, and existing options available to the analyst. A analyst may start to shut down a connection and have the AI jump in to warn about the potential impact or offer another approach that is a better option including the Jarvis like % of success the current action will have.
Keep an eye out on AI agents within the security space ….