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.

Endpoint

7/24/2017
10:30 AM
Inbar Raz
Inbar Raz
Commentary
Connect Directly
Twitter
LinkedIn
RSS
E-Mail vvv
50%
50%

Bots Make Lousy Dates, But Not Cheap Ones

The danger of dating sites: If a beautiful woman asks men to click on malware, they'll probably click.

Bill installed the dating app on his smartphone. To his surprise, he was quickly matched up with several women he found attractive. Better yet, they immediately showed their interest by sending him text messages. 

"One's a flight attendant and three are models!" he told his friends over coffee. "Why didn't I jump into online dating years ago?"

Bill's favorite four ladies all sent him numerous, nearly identical text messages that were inviting yet vague — they didn't respond directly to his questions. Bill didn't care. They were gorgeous and texting with him. He noticed that three had attended the same college, and two worked for the same agency in London. Within a day, each sent him a link and wrote Bill, "Want to meet? Check this out."

I Never Do This, But…Click!
Bill clicked on the links, and so do 70% of men when they unexpectedly receive a link sent by a bot posing as an attractive woman. That's what PerimeterX observed when we researched top dating sites. This 70% click-through rate may well be the highest conversion rate in the world, and it explains why dating bots pay off for hackers.

One of the first clues to the prevalence of bots on dating sites such as Tinder: Many female profiles in specific cities (Copenhagen and Denver, for example) list similar jobs — modeling or flight attendants, usually — and supposedly attended one of a limited set of colleges. Oddly, they often list workplaces in other cities. Invariably, their photos portray them as above average in appearance. 

Why Bots Want to "Date" You
Dating bots are outgoing and were very quick to match up with me when I posted a profile on Tinder. They also texted me and suggested that I click on links they sent. The links lure men from dating sites to porn sites, or URLs where they can be tricked into downloading malware or giving up money or personal data.

Between 22% and 35% of relationships now start online, and malicious bots are estimated to make up 29% of traffic on enterprise sites, according to a report quoted in Digital Trends. Inevitably, online dating and bots intersected. Hackers create bots on dating sites to steal traffic, obtain personal and financial data on customers, and sometimes defraud them.

We found the same bot infestations on other dating sites, so it's not only Tinder and its users that are being targeted.

Both Men and Women Exploited on Dating Sites
When we contacted a real-life user whose photographs had been hijacked to set up bot profiles, she was able to have her images removed from the site. For every bot profile, there's a face that belongs to a real person who probably is not aware her or his photos are being misused.

That's almost certainly true of the thousands of Tinder users in the Bay Area whose images were scraped and put into a public-domain facial data set. Using automated tools, scam artists copied from Tinder 20,000 profile images of women and 20,000 of men from Tinder without their knowledge.

Relatively primitive bots can make matches, start a text or email conversation, and ask men to click a link that leads to paid content sites (that is, porn). Newer, advanced bots can vary their behavior to be more convincing companions.

Men who fall for these bots may be lured into entering their credit card data on a site of ill repute or a phony "profile verification" service, and then be too embarrassed to report that they were tricked into accepting a porn site subscription.

Dating Sites Get Stood Up by Bots
Bots have a serious impact on dating sites, which lose traffic, advertising revenue, and subscription fees. A site's reputation suffers when male customers discover that a large portion of attractive women — by far the most communicative — on the site are fake. It's clear that dating bots destroy the customer experience and probably hurt customer loyalty as well. 

In addition, a dating site might be liable if a user could prove that a malware infection or fraud loss resulted from link sent to him via the site.

People Don't Spot Fake People Well
Dating sites have used humans to check new profiles. This approach is not scalable and it's unreliable. People, it turns out, aren't very effective at catching bot profiles. 

Dating sites bear the burden of stopping bot activity. It is imperative that they do the following:

  • Create a safer environment for their customers
  • Sustain a legitimate, nondeceptive customer experience
  • Protect the mission of providing an honest forum for new relationships
  • Sustain revenue and traffic

Stopping Bots Is Tough When You Can't See Them
If simpler bots are used to chat with human customers, they may be caught with more traditional defenses. Newer, more sophisticated bots are much more elusive and can be directed to vary their behavior, making them undetectable by signature-based security tools.

It's essential to avoid blocking legitimate customer traffic — that is not an acceptable by-product of protecting users and the site's reputation. Sites must be able to stop automated account creation and step in the moment a human-created profile begins to employ automated methods to communicate with unsuspecting users.

Dating sites can now turn to new behavior-based approaches to Web protection. One such method learns how human users interact with the site and spots subtle deviations from human behavior. The best practice is to check the profile of every user and all their interactions with the site, in real time. It may lead to fewer supermodels chasing after Bill and guys like him — but will also leave bots out of the dating game and help dating sites protect their reputation and users. 

Related Content:

Inbar Raz has been teaching and lecturing about Internet security and reverse engineering for nearly as long as he's been doing that himself. He started programming at the age of 9 and reverse engineering at the age of 14. Raz specializes in an outside-the-box approach to ... View Full Bio
Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
97% of Americans Can't Ace a Basic Security Test
Steve Zurier, Contributing Writer,  5/20/2019
TeamViewer Admits Breach from 2016
Dark Reading Staff 5/20/2019
How a Manufacturing Firm Recovered from a Devastating Ransomware Attack
Kelly Jackson Higgins, Executive Editor at Dark Reading,  5/20/2019
Register for Dark Reading Newsletters
White Papers
Video
Cartoon Contest
Write a Caption, Win a Starbucks Card! Click Here
Latest Comment: This comment is waiting for review by our moderators.
Current Issue
Building and Managing an IT Security Operations Program
As cyber threats grow, many organizations are building security operations centers (SOCs) to improve their defenses. In this Tech Digest you will learn tips on how to get the most out of a SOC in your organization - and what to do if you can't afford to build one.
Flash Poll
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2019-7844
PUBLISHED: 2019-05-22
Adobe Media Encoder version 13.0.2 has an out-of-bounds read vulnerability. Successful exploitation could lead to information disclosure.
CVE-2017-9809
PUBLISHED: 2019-05-22
OX Software GmbH OX App Suite 7.8.4 and earlier is affected by: Information Exposure.
CVE-2018-12886
PUBLISHED: 2019-05-22
stack_protect_prologue in cfgexpand.c and stack_protect_epilogue in function.c in GNU Compiler Collection (GCC) 4.1 through 8 (under certain circumstances) generate instruction sequences when targeting ARM targets that spill the address of the stack protector guard, which allows an attacker to bypas...
CVE-2019-7834
PUBLISHED: 2019-05-22
Adobe Acrobat and Reader versions 2019.010.20100 and earlier, 2019.010.20099 and earlier, 2017.011.30140 and earlier, 2017.011.30138 and earlier, 2015.006.30495 and earlier, and 2015.006.30493 and earlier have a use after free vulnerability. Successful exploitation could lead to arbitrary code execu...
CVE-2019-7835
PUBLISHED: 2019-05-22
Adobe Acrobat and Reader versions 2019.010.20100 and earlier, 2019.010.20099 and earlier, 2017.011.30140 and earlier version, 2017.011.30138 and earlier, 2015.006.30495 and earlier, and 2015.006.30493 and earlier have a use after free vulnerability. Successful exploitation could lead to arbitrary co...