Endpoint

8/10/2018
10:30 AM
Connect Directly
LinkedIn
RSS
E-Mail vvv
50%
50%

The Enigma of AI & Cybersecurity

We've only seen the beginning of what artificial intelligence can do for information security.

Alan Turing is famous for several reasons, one of which is that he cracked the Nazis' seemingly unbreakable Enigma machine code during World War II. Later in life, Turing also devised what would become known as the Turing test for determining whether a computer was "intelligent" — what we would now call artificial intelligence (AI). Turing believed that if a person couldn't tell the difference between a computer and a human in a conversation, then that computer was displaying AI.

AI and information security have been intertwined practically since the birth of the modern computer in the mid-20th century. For today's enterprises, the relationship can generally be broken down into three categories: incident detection, incident response, and situational awareness — i.e., helping a business understand its vulnerabilities before an incident occurs. IT infrastructure has grown so complex since Turing's era that it can be months before personnel notice an intrusion.

Current iterations of computer learning have yielded promising results. Chronicle, which was recently launched by Google's parent company, Alphabet, allows companies to tap its enormous processing power and advanced machine learning capabilities to scan IT infrastructure for unauthorized activity. AI² quickly learns how to differentiate true attacks from merely unusual activity, alleviating a vexing problem for IT security teams: false positives. There are numerous other examples of AI-based solutions, such as Palo Alto Networks' Magnifier, which uses machine learning to automate incident response, utilizing another strength of AI: speed.

These advances arrive at an opportune moment because the risks from cybercrime are rapidly growing; estimates of the cost worldwide is about $600 billion annually. The average cost of a data breach is estimated at $1.3 million for enterprises and $117,000 for small businesses, and companies are taking note. According to ESG research, 12% of enterprise organizations have already deployed AI-based security analytics extensively, and 27% have deployed AI-based security analytics on a limited basis.

Moreover, cybersecurity in the years ahead will be increasingly challenging. Enterprises and computers are relatively static and well-defined at present, but securing information amid the Internet of Things, in which almost every device will be programmable and therefore hackable, is going to be far harder. Soon, we won't just have to safeguard unseen servers anymore but also our cars and household devices.

Unfortunately, AI has also become available to hackers as well. Dark Web developments to date merit serious discussion, such as machine learning that gets better and better at phishing — tricking people into opening imposter messages in order to hack them. Further down the road, machines could take impersonation one step further by learning how to build fake images. Experts are also worried AI-based hacking programs might reroute or even crash self-piloting vehicles, such as delivery drones.

I suspect that in the future, users on the front end will be blissfully unaware that behind the scenes battles between good and bad learning machines rage, with each side continually innovating to outsmart the other. Already, the synthesis of AI and cybersecurity has yielded fascinating results, and there is no doubt we are only at the beginning. I am reminded of a quote by Dr. Turing:"We can only see a short distance ahead, but we can see plenty there that needs to be done."

Related Content:

Learn from the industry's most knowledgeable CISOs and IT security experts in a setting that is conducive to interaction and conversation. Early-bird rate ends August 31. Click for more info

Dr. Dongyan Wang is Chief AI Officer at DeepBrain Chain, the world's first AI computing platform powered by the blockchain. Dr. Wang has almost 20 years of experience in AI and data science, including at several Fortune 500 companies. Among other accomplishments, Dr. Wang has ... View Full Bio
Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
Higher Education: 15 Books to Help Cybersecurity Pros Be Better
Curtis Franklin Jr., Senior Editor at Dark Reading,  12/12/2018
Worst Password Blunders of 2018 Hit Organizations East and West
Curtis Franklin Jr., Senior Editor at Dark Reading,  12/12/2018
Register for Dark Reading Newsletters
White Papers
Video
Cartoon Contest
Current Issue
10 Best Practices That Could Reshape Your IT Security Department
This Dark Reading Tech Digest, explores ten best practices that could reshape IT security departments.
Flash Poll
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2018-20168
PUBLISHED: 2018-12-17
Google gVisor before 2018-08-22 reuses a pagetable in a different level with the paging-structure cache intact, which allows attackers to cause a denial of service ("physical address not valid" panic) via a crafted application.
CVE-2018-20167
PUBLISHED: 2018-12-17
Terminology before 1.3.1 allows Remote Code Execution because popmedia is mishandled, as demonstrated by an unsafe "cat README.md" command when \e}pn is used. A popmedia control sequence can allow the malicious execution of executable file formats registered in the X desktop share MIME typ...
CVE-2018-20161
PUBLISHED: 2018-12-15
A design flaw in the BlinkForHome (aka Blink For Home) Sync Module 2.10.4 and earlier allows attackers to disable cameras via Wi-Fi, because incident clips (triggered by the motion sensor) are not saved if the attacker's traffic (such as Dot11Deauth) successfully disconnects the Sync Module from the...
CVE-2018-20159
PUBLISHED: 2018-12-15
i-doit open 1.11.2 allows Remote Code Execution because ZIP archives are mishandled. It has an upload feature that allows an authenticated user with the administrator role to upload arbitrary files to the main website directory. Exploitation involves uploading a ".php" file within a "...
CVE-2018-20157
PUBLISHED: 2018-12-15
The data import functionality in OpenRefine through 3.1 allows an XML External Entity (XXE) attack through a crafted (zip) file, allowing attackers to read arbitrary files.