Story highlights
Darktrace uses AI to spot emerging cyber threats
Its system learns and adapts to fight attacks
Every year cyber attacks are becoming more sophisticated and frequent – and it’s predicted they’ll cost the world $8 trillion by 2022.
But a company founded by mathematicians and ex-spies is trying to change the way organizations protect themselves from hackers.
British cyber security firm Darktrace is using artificial intelligence to create a “cyber immune system.”
Usually, companies try to anticipate what an attack would look like and ask their computers to look for anomalies that match that description. But Darktrace says it has come up with a system that learns what is normal behavior within a company’s network, and can then make intelligent judgments to spot emerging threats – even if it experiences a new, complex attack that it’s never seen before.
Protecting the UK’s largest power supplier
Drax Power Station uses Darktrace’s system. It’s the largest power plant in the UK and if Drax’s systems were to be compromised, it could leave thousands of homes without electricity.
Drax’s group head of security, Martin Sloan, says before Darktrace the company had a very traditional cyber security strategy.
READ: Darkest material on Earth will create a ‘schism in space’
“You’ve got several layers of firewalls, but there is an acceptance that no matter how good those defenses are, you’ll still be breached,” he said.
‘This technology acts like a fever’
Instead of just building firewalls, the Darktrace Enterprise Immune System is designed to understand what the company’s normal network looks like and identify any abnormalities.
Sloan says the system behaves the same way as the body when it has the flu: “This technology is like a fever that alerts us when we have a virus and then we need to take action to treat it.”
READ: The tiny Welsh car that runs on hydrogen and emits only water
An example of a small abnormality that the technology would pick up is if an employee logged onto the server at 10pm, without ever have done so before. It would be immediately flagged as unusual.
Dave Palmer, director of technology at Darktrace, explains more: “We don’t have to sit there and tell it, ‘hey, this is what normal looks like’ or ‘here’s how Dave uses his laptop.’
“The system can learn all of those things on its own and build a complex pattern of life understanding purely by observing what normally happens.”
Watch the video above to find out more.
