AI-powered security for higher education IT concerns

By Trent Fierro, Blog Contributor

Finding ways to improve IT efficiency in higher education (community colleges and universities) often leads to conversations I’m having about AI network-oriented features for Wi-Fi concerns. But using AI to solve security challenges is also being discussed more frequently. This is because of a student’s ability to connect from anywhere with almost any endpoint imaginable—really old laptops and phones as well as newly released ones and IoT devices such as game consoles, smart lightbulbs, and wearables.

The self-onboarding of endpoints is a big challenge. As these devices make it onto wired and wireless networks, it’s difficult to ensure they’re given proper access privileges due to their variety and likelihood that some will be labelled “unknown”. If the endpoint doesn’t match a known profile, it’s likely that it will be denied access. A workaround has been to provide open access or use MPSK on higher ed campuses, but neither provide the visibility that security teams really want.

The mentality going forward is to use AI and automation to tackle network and security issues more cohesively as the frequency, impact, and cost of a breach can be daunting. And buying separate tools isn’t helping the bottom line.

In a presentation by Bob Laliberte, Principal Analyst with ESG, that I mentioned in a separate blog on AI for network aspects, his data reinforced what I was hearing from IT teams about AI helping with security concerns as well. Members of the TechTarget network in higher ed are researching network AIOps and automation content with repeated frequency.

ESG TechTarget Who is researching network automation and AIOps

What this means is that IT staff and leaders understand that AI and automation can and will help, solutions are available today, and that AIOps doesn’t take away from their roles—it adds to it.

From a security perspective, Aruba Central with built-in AIOps includes automated Client Insights, an AI-powered device profiling solution that uses very intricate machine learning models to pinpoint the manufacturer of an endpoint client. Client Insights also shows you the normal behavior of an endpoint and any abnormal behavior.

Here are a few ways to use Client Insights:

  • Imagine using just one system (Aruba Central) to write rules that can be enforced using highly accurate endpoint types and roles—student-owned computers and IoT versus facilities-owned IoT for instance. No other vendor provides the ability to use their cloud management solution to do this.
  • Automatically assign “group” access privileges and segment per location, SSID, port, or other attributes to eliminate manual steps per user or endpoint being used, especially if MPSK is being used.
  • Identify if an endpoint is behaving differently is just a few clicks. For instance, it’s easy to see if an endpoint is reaching out to websites that are not normal or roaming when you know it’s a stationary device like a large printer. And then, trigger a change of authorization.

As you explore how AI and automation can help on your campus, remember that a unified approach is important. You want AIOps for network and security use cases, not just one piece of the proverbial pie. All of your access points and switches help Aruba Central with automated Client Insights and help our AI engine understand what’s being connected, from where and how they’re behaving—all starting with basic Foundation Licensing and no dedicated collectors needed.

Here are some additional resources that you may find useful: