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.

Operations

4/18/2016
08:00 AM
Connect Directly
Twitter
RSS
E-Mail
50%
50%

MIT AI Researchers Make Breakthrough On Threat Detection

New artificial intelligence platform offers 3x detection capabilities with 5x fewer false positives.

Researchers with MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) believe that can offer the security world a huge boost in incident response and preparation with a new artificial-intelligence platform it believes can eventually become a secret weapon in squeezing the most productivity from security analyst teams.

Dubbed AI2, the technology has shown the capability to offer three times more predictive capabilities and drastically fewer false positive than todays analytics methods.

CSAIL gave a sneak peek into AI2 in a presentation to the academic community last week at the IEEE International Conference on Big Data Security, which detailed the specifics of a paper released to the public this morning. The driving force behind AI2 is its blending of artificial intelligence with what researchers at CSAIL call "analyst intuition," essentially finding an effective way to continuously model data with unsupervised machine learning while layering in periodic human feedback from skilled analysts to inform a supervised learning model.

"You can think about the system as a virtual analyst,” says CSAIL research scientist Kalyan Veeramachaneni, who developed AI2 with former CSAIL postdoc Ignacio Arnaldo, who is now a chief data scientist at PatternEx. “It continuously generates new models that it can refine in as little as a few hours, meaning it can improve its detection rates significantly and rapidly.”

This offers the best of both worlds in what has become a bright line division in security analytics today. For the most part, security systems today either depend on analyst-driven solutions that rely on rules created by human experts or they lean heavily on machine-learning systems for anomaly detection that trigger highly disruptive false positive rates.

Gain insight into the latest threats and emerging best practices for managing them. Attend the Security Track at Interop Las Vegas, May 2-6. Register now!

In the paper released today, Veeramachaneni, Arnaldo and their team showed how the system did when tested with 3.6 billion pieces of log data generated by millions of users over three months. During this test, the platform was able to detect 85% of attacks, three times better than previous benchmark, while at the same time reducing false positives by a factor of five.

The approach of melding together human- and computer-based approaches to machine learning has long run into stumbling blocks due to the challenge of manually labeling cybersecurity data for algorithms. The specialized nature of analyzing the data makes it a difficult data set to crack with typical crowdsourcing strategies employed in other arenas of big data analysis. The average person on a site like Amazon Mechanical Turk would be hard-pressed to apply accurate labels for data indicating DDoS or exfiltration attacks, Veermachaneni explained.

Meanwhile, security experts have already tried several generations worth of supervised machine learning models only to find that 'feeding' these systems ends up creating more work rather than saving an analyst time. This is what has lead many organization to dump early analytics solutions in the proverbial waste bin after experiencing those frustrations.

AI2 is able to perform better by bringing together three different unsupervised learning models to sift through raw data before presenting data to the analyst. So on day one, that system offers 200 of the most abnormal events to an analyst, who then manually sifts through those to identify the real attacks. That information is fed back into the system and even within a few days the unsupervised system is presenting as few as 30 to 40 events for verification.

“The more attacks the system detects, the more analyst feedback it receives, which, in turn, improves the accuracy of future predictions,” Veeramachaneni says. “That human-machine interaction creates a beautiful, cascading effect.”

Check out this video for a quick overview of the way AI2 works.

Related Content: 

 

Ericka Chickowski specializes in coverage of information technology and business innovation. She has focused on information security for the better part of a decade and regularly writes about the security industry as a contributor to Dark Reading.  View Full Bio

Comment  | 
Print  | 
More Insights
Comments
Threaded  |  Newest First  |  Oldest First
ONIHadoop
100%
0%
ONIHadoop,
User Rank: Apprentice
4/19/2016 | 5:46:28 PM
Open Network Insight
This is very interesting work, thanks for reporting on it.  We agree that there needs to be a new approach to cyber security that leads with machine learning against large datasets.  Our open source project, Open Network Insight, was launched on this same premise and provides insight and operational analytics for network flows, domain name service data and full packet captures.  

If anyone reading this article is interested in learning more please visit our website or github.
Mike Anders
50%
50%
Mike Anders,
User Rank: Apprentice
6/7/2016 | 12:24:22 AM
Re: Open Network Insight
Object Based Production (OBP) and Activity Based Intelligence (ABI) can assist in achieving "the best of both worlds" with respect to detection when both OBP and ABI are brought to bear on the cyber problem. Not a lecture, just an observation!
Data Leak Week: Billions of Sensitive Files Exposed Online
Kelly Jackson Higgins, Executive Editor at Dark Reading,  12/10/2019
Intel Issues Fix for 'Plundervolt' SGX Flaw
Kelly Jackson Higgins, Executive Editor at Dark Reading,  12/11/2019
Register for Dark Reading Newsletters
White Papers
Video
Cartoon
Current Issue
The Year in Security: 2019
This Tech Digest provides a wrap up and overview of the year's top cybersecurity news stories. It was a year of new twists on old threats, with fears of another WannaCry-type worm and of a possible botnet army of Wi-Fi routers. But 2019 also underscored the risk of firmware and trusted security tools harboring dangerous holes that cybercriminals and nation-state hackers could readily abuse. Read more.
Flash Poll
Rethinking Enterprise Data Defense
Rethinking Enterprise Data Defense
Frustrated with recurring intrusions and breaches, cybersecurity professionals are questioning some of the industrys conventional wisdom. Heres a look at what theyre thinking about.
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2019-5252
PUBLISHED: 2019-12-14
There is an improper authentication vulnerability in Huawei smartphones (Y9, Honor 8X, Honor 9 Lite, Honor 9i, Y6 Pro). The applock does not perform a sufficient authentication in a rare condition. Successful exploit could allow the attacker to use the application locked by applock in an instant.
CVE-2019-5235
PUBLISHED: 2019-12-14
Some Huawei smart phones have a null pointer dereference vulnerability. An attacker crafts specific packets and sends to the affected product to exploit this vulnerability. Successful exploitation may cause the affected phone to be abnormal.
CVE-2019-5264
PUBLISHED: 2019-12-13
There is an information disclosure vulnerability in certain Huawei smartphones (Mate 10;Mate 10 Pro;Honor V10;Changxiang 7S;P-smart;Changxiang 8 Plus;Y9 2018;Honor 9 Lite;Honor 9i;Mate 9). The software does not properly handle certain information of applications locked by applock in a rare condition...
CVE-2019-5277
PUBLISHED: 2019-12-13
Huawei CloudUSM-EUA V600R006C10;V600R019C00 have an information leak vulnerability. Due to improper configuration, the attacker may cause information leak by successful exploitation.
CVE-2019-5254
PUBLISHED: 2019-12-13
Certain Huawei products (AP2000;IPS Module;NGFW Module;NIP6300;NIP6600;NIP6800;S5700;SVN5600;SVN5800;SVN5800-C;SeMG9811;Secospace AntiDDoS8000;Secospace USG6300;Secospace USG6500;Secospace USG6600;USG6000V;eSpace U1981) have an out-of-bounds read vulnerability. An attacker who logs in to the board m...