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.

Cloud

3/13/2020
12:35 PM
Connect Directly
Twitter
RSS
E-Mail

What Cybersecurity Pros Really Think About Artificial Intelligence

While there's a ton of unbounded optimism from vendor marketing and consultant types, practitioners are still reserving a lot of judgment.
4 of 7

Limitations of Cybersecurity AI
A study by Osterman Research showed that part of the problem in this early stage of deployment is a strong perception that AI isn't quite ready for showtime yet. Some of the common complaints included issues with inaccurate results, the performance trade-offs of placing certain types of AI platforms on the endpoint, difficulty of use, and the ever-present concern over false-positives.
Image Source: 'The State of AI in Cybersecurity: The Benefits, Limitations and Evolving Questions,' Osterman Research on behalf of ProtectWise

Limitations of Cybersecurity AI

A study by Osterman Research showed that part of the problem in this early stage of deployment is a strong perception that AI isn't quite ready for showtime yet. Some of the common complaints included issues with inaccurate results, the performance trade-offs of placing certain types of AI platforms on the endpoint, difficulty of use, and the ever-present concern over false-positives.

Image Source: "The State of AI in Cybersecurity: The Benefits, Limitations and Evolving Questions," Osterman Research on behalf of ProtectWise

4 of 7
Comment  | 
Print  | 
Comments
Newest First  |  Oldest First  |  Threaded View
rstatsinger
50%
50%
rstatsinger,
User Rank: Apprentice
4/18/2020 | 5:20:24 PM
AI sweetspot: lots of date
Remember that AI thrives on large datasets against which the AI can be trained - or can self-organize - to detect patterns and draw correct inferences. This implies that the most natural fit for AI in cybersecurity is in perimeter defense - WAFs, NGWAFs, IDSs, etc - which are constantly bombarded with traffic from both the 'good guys' and the 'bad guys'. Using AI to help distinguish which is which - and to present humans with business decision assistance - is probably the lowest hanging fruit for AI in our business.
Waltsz61@gmail.com
50%
50%
[email protected],
User Rank: Apprentice
4/15/2020 | 10:22:02 AM
AI needs network and detailed end point dat
AI will always work if given the raw network and endpoint data that the process needs.  Without data, AI can not work.

99% of organizations have no idea of what's on their network, that includes users, endpoints, software and hardware and end point behavior.

what do you expect AI to analyze, you might as well use a crystal ball.  That works without data and your results will be better.

Walt

cleararmor.com

9083105916

 
Edge-DRsplash-10-edge-articles
7 Old IT Things Every New InfoSec Pro Should Know
Joan Goodchild, Staff Editor,  4/20/2021
News
Cloud-Native Businesses Struggle With Security
Robert Lemos, Contributing Writer,  5/6/2021
Commentary
Defending Against Web Scraping Attacks
Rob Simon, Principal Security Consultant at TrustedSec,  5/7/2021
Register for Dark Reading Newsletters
White Papers
Video
Cartoon
Current Issue
2021 Top Enterprise IT Trends
We've identified the key trends that are poised to impact the IT landscape in 2021. Find out why they're important and how they will affect you today!
Flash Poll
How Enterprises are Developing Secure Applications
How Enterprises are Developing Secure Applications
Recent breaches of third-party apps are driving many organizations to think harder about the security of their off-the-shelf software as they continue to move left in secure software development practices.
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2021-32606
PUBLISHED: 2021-05-11
In the Linux kernel 5.11 through 5.12.2, isotp_setsockopt in net/can/isotp.c allows privilege escalation to root by leveraging a use-after-free. (This does not affect earlier versions that lack CAN ISOTP SF_BROADCAST support.)
CVE-2021-3504
PUBLISHED: 2021-05-11
A flaw was found in the hivex library in versions before 1.3.20. It is caused due to a lack of bounds check within the hivex_open function. An attacker could input a specially crafted Windows Registry (hive) file which would cause hivex to read memory beyond its normal bounds or cause the program to...
CVE-2021-20309
PUBLISHED: 2021-05-11
A flaw was found in ImageMagick in versions before 7.0.11 and before 6.9.12, where a division by zero in WaveImage() of MagickCore/visual-effects.c may trigger undefined behavior via a crafted image file submitted to an application using ImageMagick. The highest threat from this vulnerability is to ...
CVE-2021-20310
PUBLISHED: 2021-05-11
A flaw was found in ImageMagick in versions before 7.0.11, where a division by zero ConvertXYZToJzazbz() of MagickCore/colorspace.c may trigger undefined behavior via a crafted image file that is submitted by an attacker and processed by an application using ImageMagick. The highest threat from this...
CVE-2021-20311
PUBLISHED: 2021-05-11
A flaw was found in ImageMagick in versions before 7.0.11, where a division by zero in sRGBTransformImage() in the MagickCore/colorspace.c may trigger undefined behavior via a crafted image file that is submitted by an attacker processed by an application using ImageMagick. The highest threat from t...