Dark Reading is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them.Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Operational Security //

AI

4/2/2018
09:35 AM
Simon Marshall
Simon Marshall
Simon Marshall
50%
50%

Red Bull Powers Security Strategy With AI, Automation

When it comes to security, Red Bull is looking to close the gap by turning toward newer technologies, including automation, AI and machine learning.

Red Bull is well-known for projecting an energetic brand. Behind the scenes, its IT security team also likes to be energetic, however, not in the way the company's commercials would have you believe.

Despite the play-hard image of the brand, the Red Bull team likes to be very Zen. About a year ago, it began investing in automating some of its security processes, so the organization could free-up detection and response resources to become higher value, less tactical brains.

At times, an enterprise security strategy can be dangerous when it gets overly defensive. However, when security teams want to be strategic, as Red Bull has shown, automation technology can actually help the security team think, and not just act. (See Unknown Document 736875)

"We don't want to lose the right focus or become over-protective," Jimmy Heschl, Red Bull's CISO, told Security Now, explaining how sometimes reacting to and resolving an incident can be a mistake. Even reacting and remediating correctly, shouldn't ideally -- in his world -- be done manually because it's at the cost of contending against hackers who have time on their hands and are very inventive.

"Overwhelming or excessively intrusive security controls are significant roadblocks, when [we] want to be creative, spontaneous and innovative," Heschl said. "Overreaction from security -- as this is done by colleagues that are primarily driven by various compliance requirements -- has a significant impact on these objectives."

Advent of security automation
A number of tech vendors including Demisto, IBM's Resilient Systems, Microsoft's Hexadite, and Red Bull's vendor, EnSilo, are capturing the mood with orchestration and automation offerings, powered by artificial intelligence, and more specifically, machine learning. (See Automation Answers Security Skills Shortage.)

Gartner's 2017 "Innovation Insight for Security Orchestration, Automation and Response" report finds enterprises hobbled because of analyst time lost to manual, heavy-lift processes.

"Security operations still primarily rely on manually created and maintained, document-based procedures for operations, which leads to issues such as longer analyst onboarding times, stale procedures, tribal knowledge and inconsistencies in executing operational functions," according to the report.

Increasingly, the engine behind endpoint detection and response (EDR) system automation is AI and machine learning. These technologies are in the hype curve and for some organizations, offer not only to automate manual work, but to actively couple learned threat knowledge with their own business security policies and then independently remediate attacks.

But a lack of human intervention, on the other hand, worries Red Bull, for one.

"Automated response is a challenge in itself," Heschl said. "It has to do with giving away control, and automation always has some drawbacks. It's not the detection function that I fear, but automated response from simple mail filters and network blocks; via user and access management to advanced countermeasures: the more complexity you have in response, the more that can go wrong."

The cost of dwell time
The elapsed time between threat detection and response -- dwell time -- is what costs enterprises money in terms of increased risk of data theft or damage, and the price of running through investigation and remediation processes that usually take months.

A 2017 study by the Ponemon Institute of 419 companies, entitled "The Cost of Data Breach," reported that the time to identify and the time to contain malicious attacks were an average 214 and 77 days respectively. The average cost per breach is currently about $4 million.


Want to hear more about the leading operator use cases for AI technologies? Join us in Austin from May 14-16 at the fifth-annual Big Communications Event. There's still time to register and communications service providers get in free!

Although a current drive towards zero dwell time is noble, it's a massive challenge. Fortunately, a more realistic return on investment in automated EDR is already benefiting Red Bull.

"It's the speed of initiating action [that's important]," Heschl said. "On the other side, it's the automation of response that leaves [us] independent of scarce resources.

"It helps me address my big fear: losing focus. My team can use their time to think and to improve rather than hunt adversaries," Heschl added.

Although it's the computing and learning crunch power that AI and machine learning that support this drive, despite the hype, the technology itself is relatively unimportant.

"I believe that machine learning and AI are the means to meet and achieve security initiatives," EnSilo CEO Roy Katmor said. "[But] organizations believe in added value -- namely alert efficacy, in pre- and post-infection, and operational efficiency via automation. The technology behind it is less relevant."

Related posts:

— Simon Marshall, Technology Journalist, special to Security Now

Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
More SolarWinds Attack Details Emerge
Kelly Jackson Higgins, Executive Editor at Dark Reading,  1/12/2021
Vulnerability Management Has a Data Problem
Tal Morgenstern, Co-Founder & Chief Product Officer, Vulcan Cyber,  1/14/2021
Register for Dark Reading Newsletters
White Papers
Video
Cartoon
Current Issue
2020: The Year in Security
Download this Tech Digest for a look at the biggest security stories that - so far - have shaped a very strange and stressful year.
Flash Poll
Assessing Cybersecurity Risk in Today's Enterprises
Assessing Cybersecurity Risk in Today's Enterprises
COVID-19 has created a new IT paradigm in the enterprise -- and a new level of cybersecurity risk. This report offers a look at how enterprises are assessing and managing cyber-risk under the new normal.
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2020-36192
PUBLISHED: 2021-01-18
An issue was discovered in the Source Integration plugin before 2.4.1 for MantisBT. An attacker can gain access to the Summary field of private Issues (either marked as Private, or part of a private Project), if they are attached to an existing Changeset. The information is visible on the view.php p...
CVE-2020-36193
PUBLISHED: 2021-01-18
Tar.php in Archive_Tar through 1.4.11 allows write operations with Directory Traversal due to inadequate checking of symbolic links, a related issue to CVE-2020-28948.
CVE-2020-7343
PUBLISHED: 2021-01-18
Missing Authorization vulnerability in McAfee Agent (MA) for Windows prior to 5.7.1 allows local users to block McAfee product updates by manipulating a directory used by MA for temporary files. The product would continue to function with out-of-date detection files.
CVE-2020-28476
PUBLISHED: 2021-01-18
All versions of package tornado are vulnerable to Web Cache Poisoning by using a vector called parameter cloaking. When the attacker can separate query parameters using a semicolon (;), they can cause a difference in the interpretation of the request between the proxy (running with default configura...
CVE-2020-28473
PUBLISHED: 2021-01-18
The package bottle from 0 and before 0.12.19 are vulnerable to Web Cache Poisoning by using a vector called parameter cloaking. When the attacker can separate query parameters using a semicolon (;), they can cause a difference in the interpretation of the request between the proxy (running with defa...