Machine learning can help detect unusual patterns in digital communications

Cisco’s annual Cyber Security Report says a rapidly growing number of companies suffer cyber attacks and many are resorting to new technologies to defend themselves.

Roughly one-third of all cyber security products and services are using artificial intelligence or machine learning, according to the latest issue of the report by the networking group.

Today already, 39 pc of companies use automated software, while 34 pc deploy machine learning and 32 pc use artificial intelligence to more quickly identify threats and take protective measures.

“Developments in malware in recent months show that attackers are learning and are getting increasingly astute in exploiting security loopholes,” Klaus Lenssen, chief security officer at Cisco, said in a company blog.

That means corporate IT departments have to act. “They have to optimize their security strategy, automate processes and invest in new technologies to keep pace with the attackers,” Lenssen said.

Machine learning can help software recognize unusual communications, he said.

Cisco said security threats are often underestimated in the Internet of Things (IoT). Companies are aware of the rising tide of so-called distributed denial of service (DDOS) attacks, but only 13 percent of executives polled were worried about cyber attacks taking place through IoT botnets.

The complexity of today’s corporate security landscape also poses a particular challenge, Cisco said.