Internet HoneyGrid reveals 95% of User Generated Content is spam or malicious

happygeek 1 Tallied Votes 2K Views Share

Websense Security Labs has its bi-annual State of Internet Security report and, as usual, it makes for pretty interesting if somewhat scary reading.

Covering the last six months of 2009, the report is based upon the findings of the ThreatSeeker Network which is used to discover, classify and monitor global Internet threats and trends courtesy of something called the Internet HoneyGrid. This comprises of honeyclients and honeypots, reputation systems and advanced grid computing systems, all of which combine to parse through one billion pieces of content every day while searching for security threats. Every single hour the Internet HoneyGrid scans some 40 million websites for malicious code as well as 10 million emails for unwanted content and malicious code.

So what did the HoneyGrid have to report about the Internet security threatscape for Q3/Q4 2009?

Here are the key findings:

  • 13.7% of searches for trending news/buzz words (as defined by Yahoo Buzz & Google Trends) led to malware.
  • The second half of 2009 revealed a 3.3% decline in the growth of malicious Web sites compared to the first half of the year. Websense Security Labs believes this is due to the increased focus on Web 2.0 properties with higher traffic and multiple pages.
  • However, comparing the second half of 2009 with the same period in 2008, Websense Security labs saw an average of 225% growth in malicious Web sites.
  • 71% of Web sites with malicious code are legitimate sites that have been compromised.
  • 95% of user-generated posts on Web sites are spam or malicious.
  • Consistent with previous years, 51% of malware still connects to host Web sites registered in the United States.
  • China remains second most popular malware hosting country with 17%, but during the last six months Spain jumped into the third place with 15.7% despite never having been in the top 5 countries before.
  • 81% of emails during the second half of the year contained a malicious link.
  • Websense Security Labs identified that 85.8% of all emails were spam.
  • Statistics for the second half of 2009 show spam emails broke down as 72% (HTML), 11.2% (image), 14.4% (plain text with URL) and 2.4% (plain text with no URL).
  • 35% of malicious Web-based attacks included data-stealing code.
  • 58% of all data-stealing attacks are conducted over the Web.
Ancient Dragon commented: Nice news article :) +26

Dani AI

Generated

This thread highlights a widely circulated Websense finding from early 2010 that drew strong reactions — from sharing the report to asking what “user‑generated” really meant and calling the figure into question. Coverage at the time summarized the headline and the data source; that reporting is useful context for the discussion. (arstechnica.com)

Headline numbers can be misleading unless you read the methodology. Networks that use honeyclients, honeypots or large classifier engines tend to see a disproportionate share of automated or injected content because those systems intentionally attract abuse and because classification rules (what counts as “malicious”) vary. That makes number comparisons between a global honey‑grid and a single forum or blog misleading unless you account for sampling and labelling biases. ’s point about scripted posts is spot on; they are exactly the kind of traffic these systems capture. (honeynet.onofri.org)

Practical takeaway for site owners and moderators: assume UGC will be abused and use layered defenses — rate limits, email verification, simple heuristics (hold posts with multiple links), reputation scoring, and bot checks such as reCAPTCHA. Log and measure false positives so filters stay effective rather than just noisy. For a very small, low‑friction server check, flag comments with many links, for example:

if (preg_match_all('/https?:\\/\\//i', $comment, $m) > 2) {
    // route to moderation queue
}

For broader guidance on automated threats and mitigations consult the OWASP automated‑threats material and standard bot/CAPTCHA services. (owasp.org)

Numbers like the one quoted were a wake‑up call in 2010; treat them as an alarm to review defenses, but not as a direct measure of any single community’s day‑to‑day post quality.

happygeek 2,411 Most Valuable Poster Team Colleague Featured Poster

Typical, first comment on the story is spam :)

deadcrow 0 Newbie Poster

What is meaning of:
95% of user-generated posts on Web sites are spam or malicious

If the computer creating the post is compromised it is not user generated.
If the computer creating the post is not compromised, how is the post spam or malicious?

bbronson 0 Newbie Poster

Have you not seen forums in which fake posts are made by a script advertising the latest deals in erectile dysfunction drugs? That's what it meant by 'user-generated'. Fairly obvious to most, I would think.

happygeek 2,411 Most Valuable Poster Team Colleague Featured Poster

Yep, user generated content as in forum postings, blog comments and the like. We all know that there is a certain amount of spam, and indeed malicious link spam at that, amongst UGC (our moderator team spends a great deal of time here at DaniWeb clearing the spam out of the forums) but it came as quite a surprise to see that 95% figure being quoted in the report.

deadcrow 0 Newbie Poster

Wow.. I thought most of those posts were from compromised machines and not UGC!!!

Stunning and Sad all at once that so many idiots are running around!

jarit@xs4all.nl 0 Newbie Poster

And when mixing this fake 'user generated' content with Google AdWords, both Google and the person posting the content are laughing all the way to the bank at the expense of the advertiser.

SNIP

jennypitts 3 Junior Poster

The problem with China and other countries doing malware spamming Illegal activities is that it is almost impossible to stop them.

janklaza 0 Newbie Poster

95% of user-generated posts on Web sites are spam or malicious.

Total BS on that one.

Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.