Trend Micro has given an overview of its Deep Discovery environment for Australian customers at its Evolve 2013 security conference in Sydney.
First launched internationally during 2012, Deep Discovery attracted headlines during March when the company announced it had identified the devastating attacks launched against South Korean banks by that country's rivals across the border.
Deep Discovery particularly targets the Advanced Persistent Threat – APT – attack style, which often succeed because attackers are taking care to avoid doing anything that triggers anti-malware companies' signature gathering and reporting.
CTO Raimund Genes said the service suite first arose out of failed experiments – an attempt to blow attack detection onto silicon (“we spent $50 million on the FPGA – nobody wanted it because it was too expensive”) which it later offered as a standalone software product (“which produced too many false positives”).
It wasn't until 2012 that Deep Discovery was launched as a service offering instead, with the aim of providing a better defence against the increasingly-intractable problem of advanced persistent attacks.
No matter what warnings are given about not clicking on links, social engineering remains a straightforward, simple, and successful means to get beyond corporate security defences (as was noted separately by the company's John Oliver, the malware community is devoting considerable attention to the mundane business of measuring the success of its approaches, including what kinds of e-mails are most likely to receive the desired response).
Moreover, the transition to “click this link” instead of sending attachments means there's far less forensic footprint left behind by an attacker.
Spotting those attacks when a staff member has failed the IQ test and clicked on a link, however, remains difficult. In Deep Discovery, Genes explained, the company has taken he principles it had in mind when it first began the research for its unsuccessful FPGA product, and applied them to the business of detecting targeted attacks.
“The aim is to detect, analyse, adapt and respond,” he said. To do that, it's necessary to accomplish tasks such as detecting both malicious content and malicious communications. In detecting when malware is trying to make contact with its command and control server – or when it's trying to send stolen credentials or data out of the network – the problem is that malware authors have learned to try to avoid detection, by encrypting what they send, and by staying “below the radar” of firewalls (such as by avoiding causing congestion on the outbound connection).
“Deep Discovery provides the network wide visibility and control needed to address targeted attacks,” Genes said.
To detect suspect communications, he said, the Deep Discovery environment will look at attempts to establish outbound encrypted tunnels, attempts to connect known command and control servers, and both black- and white-listing of IP address blocks.
It's worth noting that market success with Deep Discovery should theoretically help improve the service, by expanding Trend's understanding of the behaviour of Advanced Persistent Attacks and targeted attacks.
Other Evolve Security Conference Coverage for 2013: