Endpoint

1/20/2017
01:12 PM
Connect Directly
Twitter
LinkedIn
Google+
RSS
E-Mail
50%
50%

Security Chatbots Aim To Simplify Incident Detection And Response

Emerging intelligent assistants help analyze and automate response to attacks.

Security analysts soon may get extra help from intelligent assistants that help automate incident detection and response across endpoints.

Firms including Endgame and Demisto are building security chatbots to address the lack of resources and automated solutions available to today's security analysts. 

"Security teams are faced with two major challenges: insufficient resources to stop attacks in-progress and lack of automated solutions to uncover malicious behavior in time to prevent information theft," said Endgame CTO Jamie Butler in a statement.

Endgame today rolled out Artemis, an intelligent assistant integrated with its endpoint detection and response (EDR) platform. Artemis detects malicious behavior and pulls from user data to suggest responses and ultimately help save time.

Security analysts interact with Artemis via chat interface; communication is typed (not spoken). They can ask questions like "What anomalies were seen in the past 12 hours?" Artemis will present any relevant data and based on that, analysts may ask what can be done. The assistant will offer recommendations to address the problem before it escalates.

One of the goals behind Artemis was to help businesses struggling with a lack of security talent, explains Endgame data scientist Bobby Filar. Customers had expressed frustration with their lack of employees to comb through data and decide how threats should be addressed.

"We see that a lot with organizations setting up their own SOC teams or threat hunting teams," says Filar. "There just aren't enough resources to do that effectively."

Tier-one analysts don't need to understand another query language or how the data is structured, he continues. The tool helps them find anomalies, conduct investigations, and eliminate threats akin to a tier-three analyst.

Endgame consulted a range of security pros including threat hunters, forensics experts, and SOC analysts during development, adds data scientist Richard Seymour. They helped formulate questions that analysts should ask while searching for threats.

Artemis was built with natural language processing and machine learning so it will eventually be able to anticipate priorities and recognize patterns. As people start to ask questions, the tool will understand their individual habits and be able to better formulate responses, says Filar. Analysts can also write their own workflow scripts to customize it over time.

Both data scientists acknowledge Artemis has room for improvement, however. Right now the greatest shortcoming is a lack of use cases, says Filar. He hopes the tool will get better at understanding typos and diversity in vocabulary; for example, how businesses refer to different events.

Artemis is not the first nor the only chatbot for endpoint security. Demisto offers a similar tool called Demisto Enterprise, an automation and ChatOps platform for SOCs. It was recently adopted by mapping platform Esri.

ChatOps, like Artemis, is powered by intelligence. It's also designed to automate security operations and incident management so analysts can cut down on investigation and response time.

The tool is powered by DBot, a security chatbot that communicates with teams via ChatOps in the Demisto Enterprise platform. Analysts can share information, order DBot to perform actions, and learn the results it reports back.

ChatOps was created upon learning large enterprise customers struggled with documenting and sharing their actions with their teams, explains Demisto co-founder and marketing vice president Rishi Bhargava. The tool integrates with about 100 security products; analysts can use these integrations to build common workflows, which save time when alerts occur.

This tool is different from Artemis in the sense that it pulls data from a range of products, relying on them for information, he says. "It's really a security orchestration layer; it's not really a security product by itself," he explains.

Demisto's tool was also built to address the talent gap, Bhargava notes. Analysts in the same chatroom can communicate across geographies, which is necessary for teams who hire remote analysts because they can't find talent nearby. The bot learns over time to identify common actions that analysts learn. New and future analysts can be instructed on these actions, helping them develop expertise.

In the future, Demisto plans to evolve its tool to give a more analytical and deconstructed view of data to analysts, says Bhargava. It also plans to expand the amount of integrations available.

Related Content:

Kelly Sheridan is the Staff Editor at Dark Reading, where she focuses on cybersecurity news and analysis. She is a business technology journalist who previously reported for InformationWeek, where she covered Microsoft, and Insurance & Technology, where she covered financial ... View Full Bio

Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
AshishJ981
50%
50%
AshishJ981,
User Rank: Apprentice
1/19/2018 | 6:22:13 AM
End-to-end chatbot security

The question also comes that can platforms that integrate with chatbot platforms provide end to end encryption? And if not, then how is this communication channel secure (given that it doesn't have end-to-end encryption. 

Ashish Jain
www.engati.com

New Free Tool Scans for Chrome Extension Safety
Dark Reading Staff 2/21/2019
Making the Case for a Cybersecurity Moon Shot
Adam Shostack, Consultant, Entrepreneur, Technologist, Game Designer,  2/19/2019
Register for Dark Reading Newsletters
White Papers
Video
Cartoon
Current Issue
5 Emerging Cyber Threats to Watch for in 2019
Online attackers are constantly developing new, innovative ways to break into the enterprise. This Dark Reading Tech Digest gives an in-depth look at five emerging attack trends and exploits your security team should look out for, along with helpful recommendations on how you can prevent your organization from falling victim.
Flash Poll
How Enterprises Are Attacking the Cybersecurity Problem
How Enterprises Are Attacking the Cybersecurity Problem
Data breach fears and the need to comply with regulations such as GDPR are two major drivers increased spending on security products and technologies. But other factors are contributing to the trend as well. Find out more about how enterprises are attacking the cybersecurity problem by reading our report today.
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2019-6485
PUBLISHED: 2019-02-22
Citrix NetScaler Gateway 12.1 before build 50.31, 12.0 before build 60.9, 11.1 before build 60.14, 11.0 before build 72.17, and 10.5 before build 69.5 and Application Delivery Controller (ADC) 12.1 before build 50.31, 12.0 before build 60.9, 11.1 before build 60.14, 11.0 before build 72.17, and 10.5...
CVE-2019-9020
PUBLISHED: 2019-02-22
An issue was discovered in PHP before 5.6.40, 7.x before 7.1.26, 7.2.x before 7.2.14, and 7.3.x before 7.3.1. Invalid input to the function xmlrpc_decode() can lead to an invalid memory access (heap out of bounds read or read after free). This is related to xml_elem_parse_buf in ext/xmlrpc/libxmlrpc...
CVE-2019-9021
PUBLISHED: 2019-02-22
An issue was discovered in PHP before 5.6.40, 7.x before 7.1.26, 7.2.x before 7.2.14, and 7.3.x before 7.3.1. A heap-based buffer over-read in PHAR reading functions in the PHAR extension may allow an attacker to read allocated or unallocated memory past the actual data when trying to parse the file...
CVE-2019-9022
PUBLISHED: 2019-02-22
An issue was discovered in PHP 7.x before 7.1.26, 7.2.x before 7.2.14, and 7.3.x before 7.3.2. dns_get_record misparses a DNS response, which can allow a hostile DNS server to cause PHP to misuse memcpy, leading to read operations going past the buffer allocated for DNS data. This affects php_parser...
CVE-2019-9023
PUBLISHED: 2019-02-22
An issue was discovered in PHP before 5.6.40, 7.x before 7.1.26, 7.2.x before 7.2.14, and 7.3.x before 7.3.1. A number of heap-based buffer over-read instances are present in mbstring regular expression functions when supplied with invalid multibyte data. These occur in ext/mbstring/oniguruma/regcom...