Cloud

3/27/2018
12:35 PM
Connect Directly
Google+
Twitter
RSS
E-Mail
50%
50%

Bad Bots Increasingly Hide Out in Cloud Data Centers

Humans accounted for nearly 58% of website traffic in 2017 -- the rest were bad and good bots.

Bots became a household term last year in the wake of Russian election-meddling in the US and their inordinate presence on social media platforms. The population of these malicious bots also grew by nearly 10% last year, accounting for one-fifth of all website traffic.

So-called bad bots also execute online fraud, data theft, and distributed denial-of-service attacks, and despite more awareness as well as moves by Twitter and others to purge them, they continue to dog e-commerce and evolve their tactics to evade detection, according to a new analysis of bot activity by Distil Networks that studied hundreds of billions of bad bot requests on thousands of websites.

Humans accounted for nearly 58% of website traffic in 2017, with the rest bad bots (21.46%) and good bots (20.74%). Good bots include tools like search engine crawlers, while bad bots are everything from trolls to illicit data-scraping tools and proxies for cybercrime. Most bad bots live on gambling (53.1%) and airline (43.9%) websites, and most (83.2%) pose as Web browser-users, including Chrome, Firefox, Internet Explorer, and Safari, and 10.4% as mobile browsers (Safari, Android, and Opera).

The biggest shift in 2017 was bots hiding out in data centers: some 82.7% are now operating out of cloud-based accounts versus 60.1% in 2016, the data shows.

Anna Westelius, senior director of security research at Distil, says bad bots are waging credential-stuffing attacks en masse. While account takeover attacks on average occur two to three times per month, after a data breach occurs, account takeover attacks increase threefold, according to Distil's data.

"They are trying them wherever they can," Westelius says of the stolen credentials.

They're also mimicking human behavior more convincingly, by executing JavaScript like a browser, or faking mouse movements. "A lot of the time, bad bots are utilizing human connections, like human smartphone connections," Westelius says. "A lot of these are malware-related botnets" that want to appear as human as possible in their communications and behaviors, she says.

Distil found that 5.8% of all mobile devices on cellular networks are used in bad bot attacks. These bots are considered the most advanced or sophisticated because they are less likely to get detected. Overall, 74% of bad bot traffic today is sophisticated or moderately sophisticated, the report says.

But operating out of cloud data centers is all the rage for bot runners now. It's inexpensive to spin up a cloud server, for example, and it appears legit. "Hosting provides really offer them a legal way to highly distribute themselves. It's cheap and accessible," Westelius says.

The move to the cloud coincided with a decrease in residential bot traffic, according to Distil. "The economics and success of using low-cost cloud data centers probably explains why there was a drop in the amount of traffic from residential ISPs, falling from 30.5% to 14.8% in 2017," the report said.

Related Content:

 Interop ITX 2018

Join Dark Reading LIVE for an intensive Security Pro Summit at Interop ITX and learn from the industry’s most knowledgeable IT security experts. Check out the agenda here.Register with Promo Code DR200 and save $200.

Kelly Jackson Higgins is Executive Editor at DarkReading.com. She is an award-winning veteran technology and business journalist with more than two decades of experience in reporting and editing for various publications, including Network Computing, Secure Enterprise ... View Full Bio

Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
Worst Password Blunders of 2018 Hit Organizations East and West
Curtis Franklin Jr., Senior Editor at Dark Reading,  12/12/2018
8 Security Tips to Gift Your Loved Ones For the Holidays
Steve Zurier, Freelance Writer,  12/18/2018
How to Engage Your Cyber Enemies
Guy Nizan, CEO at Intsights Cyber Intelligence,  12/18/2018
Register for Dark Reading Newsletters
White Papers
Video
Cartoon Contest
Current Issue
The Year in Security 2018
This Dark Reading Tech Digest explores the biggest news stories of 2018 that shaped the cybersecurity landscape.
Flash Poll
[Sponsored Content] The State of Encryption and How to Improve It
[Sponsored Content] The State of Encryption and How to Improve It
Encryption and access controls are considered to be the ultimate safeguards to ensure the security and confidentiality of data, which is why they're mandated in so many compliance and regulatory standards. While the cybersecurity market boasts a wide variety of encryption technologies, many data breaches reveal that sensitive and personal data has often been left unencrypted and, therefore, vulnerable.
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2018-16883
PUBLISHED: 2018-12-19
sssd versions from 1.13.0 to before 2.0.0 did not properly restrict access to the infopipe according to the "allowed_uids" configuration parameter. If sensitive information were stored in the user directory, this could be inadvertently disclosed to local attackers.
CVE-2018-17192
PUBLISHED: 2018-12-19
The X-Frame-Options headers were applied inconsistently on some HTTP responses, resulting in duplicate or missing security headers. Some browsers would interpret these results incorrectly, allowing clickjacking attacks. Mitigation: The fix to consistently apply the security headers was applied on th...
CVE-2018-17193
PUBLISHED: 2018-12-19
The message-page.jsp error page used the value of the HTTP request header X-ProxyContextPath without sanitization, resulting in a reflected XSS attack. Mitigation: The fix to correctly parse and sanitize the request attribute value was applied on the Apache NiFi 1.8.0 release. Users running a prior ...
CVE-2018-17194
PUBLISHED: 2018-12-19
When a client request to a cluster node was replicated to other nodes in the cluster for verification, the Content-Length was forwarded. On a DELETE request, the body was ignored, but if the initial request had a Content-Length value other than 0, the receiving nodes would wait for the body and even...
CVE-2018-17195
PUBLISHED: 2018-12-19
The template upload API endpoint accepted requests from different domain when sent in conjunction with ARP spoofing + man in the middle (MiTM) attack, resulting in a CSRF attack. The required attack vector is complex, requiring a scenario with client certificate authentication, same subnet access, a...