Subscribe
  • Home
  • /
  • Security
  • /
  • AI fills the cyber security gaps where humans can no longer cope

AI fills the cyber security gaps where humans can no longer cope

By Tracy Burrows, ITWeb contributor.
Johannesburg, 26 Feb 2021
Mariana Pereira
Mariana Pereira

When it comes to cyber, AI is no longer a ‘nice to have’ but necessary.

This is according to Mariana Pereira, director of e-mail security products at Darktrace, who was speaking during the Securing the Future of Work with Artificial Intelligence webinar presented in partnership with ITWeb this week.

Fast-changing cyber attack modes and proliferating threats have overtaken humans’ ability to mitigate and protect their organisations against all risks. This is where artificial intelligence (AI) steps in to support security teams by monitoring and understanding all data, all processes and all activities in the enterprise environment, and automatically identifying and neutralising any anomalies. 

Cyber criminals were taking advantage of the disruption and change the world was experiencing, said Pereira. “In the past year, we saw organisations accelerating and consolidating the digital roadmaps they had planned to roll out over the next three years.” 

A lot of critical information, trade secrets and contracts reside in inboxes, on laptops and in environments where people may have a less developed awareness of the risks.

Mariana Pereira, Darktrace.

She said that in the digital collaboration space alone, there had been a 58 million seat increase in Microsoft 365, a 90% increase in mobile apps for business, a 475% increase in Microsoft Teams use, Zoom user growth of over 10 million, 20% increase in G Suite adoption and a 30% increase in Google Meet users.

“We’re not going back to the way it was before. Many changes that were made as a stop gap measure to support business continuity last year are now here to stay. Now, a lot of critical information, trade secrets and contracts reside in inboxes, on laptops and in environments where people may have a less developed awareness of the risks. This increases the risk of data loss and cyber security incidents,” she said.

 “We’ve seen that throwing humans at the growing cyber security problem doesn’t scale. In addition, humans can become tired, they can forget what happened in the past, and they can suffer from alert fatigue. AI doesn’t get tired, it doesn’t forget, and it is capable of monitoring the entire enterprise across all silos, to identify anything that doesn’t fit the pattern of normal communications and processes.”

AI works with human security teams as a force multiplier in the battle against risk and cyber crime, she said.

Darktrace is the creator of Autonomous Response technology. Its self-learning AI is modelled on the human immune system and used by over 3 000 global organisations to protect against threats to the cloud, e-mail, IOT, networks and industrial systems.

AI quickly learns what normal looks like and understands the DNA of the company.

Mariana Pereira, Darktrace.

Darktrace AI simply needs to be exposed to enterprise data, email systems and infrastructure, and it needs around 7 – 10 days to learn normal patterns within an enterprise, she said. Once the AI has learned the normal patterns in communication, processes and clusters, it works like an immune system to identify and react to any anomalies.

“AI quickly learns what normal looks like and understands the DNA of the company. In 2020 we saw that AI did very well in the transition, it learned at speed and adapted to the new way of working.

“Our AI stops an attack and empowers the organisation with information around it. It investigates the incident, rather than sending out 15 alerts which are all part of the same attack, dramatically reducing the time humans spend figuring out what is happening,” she said.

Share