How machine learning can protect your school

How machine learning can protect your school

In November, the NTT Security 2018 Global Threat Intelligence Report found that schools are just as vulnerable – if not more so – than companies.

This is because like small businesses, schools typically have little ICT know-how, and are probably at least as likely to run outdated ICT systems – especially unpatched software.

However, this is just one reason why cybersecurity will remain top of mind for schools in 2019.

Aruba’s General Manager, South Pacific, Anthony Smith, says that while many schools set up efficient firewalls and digital intrusion systems, they remain vulnerable to the most advanced persistent cyberthreats.

“Advanced persistent threats worm their way through the perimeter and make their way deep inside the school,” Smith told The Educator.

“Most schools don’t have the ability to invest in things like ICT security experts, the systems and functions that support those people and the automation and machine learning that run continuously and look for those threats.”

Smith says the “insider threat” is also worth paying close attention to.

“Defending against insider threats can also be a costly investment, and there are scarce resources in Australia to put the people and systems in place to mitigate the risks associated with this kind of threat,” Smith said.

“This is probably the next-level investment around security, protecting the inner workings of the organisation rather than just the edge of it.”

How principals can respond

Smith said schools can leverage machine learning to tackle these threats.

“There are fantastic applications for machine learning, because it can learn patterns of behaviour that it should respond to,” he said.

“It just watches the behaviour of people and devices then establishes what normal, and abnormal, looks like. This allows schools to alert their ICT department who can then fix the problem.”

Aruba’s IntroSpect User and Entity Behaviour Analytics (UEBA) uses AI-based machine learning to spot changes in user behaviour that often indicate inside attacks that have evaded perimeter defences.

Smith said this provides security teams armed with insights into malicious, compromised or negligent users, systems and devices who can cut off the threat before it does damage.

“It tells the ICT department that an abnormal event is occurring and gives them the opportunity to respond to it, rather than simply highlighting every potentially abnormal event and creating a large number of false alarms,” he said.

“People have to figure out what an event looks like and how to interpret it, but machine learning does that for you and only surfaces those issues that ICT staff really need to look at and respond to.”

Next generation wireless

Smith said the new 802.11ax (Wi-Fi 6) standard gives schools the ability to move more traffic through their Wi-Fi networks.

“The previous generations of Wi-Fi was like a round robin, but the new standard is much smarter and allows schools to pack many more devices on to a single channel. This allows schools to serve multiple clients at once,” he said.

“Students’ and staff’s phones, tablets and laptops need to have Wi-Fi 6 on them for this to work best. So K-12 schools really need to plan to go with the base foundation of this standard now, because Wi-Fi networks typically go in for five-to-ten years.”

Smith said this will allow schools to be better positioned to support the new generation products coming into the market this year and next year.

“Schools that do this will have a solid foundation for high-density networks,” he said.