Inside the Booming, Unhinged, and Dangerous Malvertising Menace

mask.of.sanity writes: The Register has a feature on the online malicious advertising (malvertising) menace that has become an explosively potent threat to end-user security on the internet. Experts say advertising networks and exchanges need to vet their customers, and publishers need to vet the third party content they display. Users should also consider script and ad blockers in the interim. From the article: "Ads as an attack vector was identified in 2007 when security responders began receiving reports of malware hitting user machines as victims viewed online advertisements. By year's end William Salusky of the SANS Internet Storms Centre had concocted a name for the attacks. Since then malvertising has exploded. This year it increased by more than 260 percent on the previous year, with some 450,000 malicious ads reported in the first six months alone, according to numbers by RiskIQ. Last year, security firm Cyphort found a 300 percent increase in malvertising. In 2013, the Online Trust Alliance logged a more than 200 percent increase in malvertising incidents compared to 2012, serving some 12.4 billion malvertisement impressions."

Comments

Advertisers, worry about security? Get real

[Anonymous Coward]

It costs money to vet customers.

For once we get to see the tragedy of the commons at work in an industry that deserves it.

Re:Advertisers, worry about security? Get real

[gweihir]

Very much so. Advertising is a plague and deserves to be eradicated. And don't tell me "it finances content", because so can crime, and apparently the distinction is not entirely clear anymore. There are other ways to finance content, and if you do not qualify, maybe your content was not valuable in the first place.

I work in online advertising

[FireballX301]

But I agree with the general premise. It's just that the picture generally gets complex - let me explain.

The way an ad gets served is this. Places that show ads (websites, mobile websites, in-app ad spaces) are inventory. Inventory is of varying quality - an ad on the front page of the NYT is costly, whereas an ad on housewiferecipes.com or something is dirt cheap. Small sites sell their inventory to brokers, who pack it up with other sites to sell on advertising exchanges (the firm I work for runs one of these exchanges).

On the other side of the issue, advertisement costs money. A firm wanting to run ads will contract with an online media agency, which will create an ad and then find inventory to place the ad in. The firm commits to spending X amount of money for Y amount of impressions (hits), so if the agency can find inventory that performs (hits whatever ad metrics required, such as 'time in ad' or 'number of clicks') while being dirt cheap, it pockets the rest. If multiple agencies bid on the same inventory, the price of that inventory goes up (and the website runner makes more money), so it's a game of scooping up cheap inventory on random sites at the times they're cheap.

Typically, a given source of inventory (a site) will contract out to a large number of brokers in order to guarantee that at least one of them will, upon request, be able to serve an ad in the space. 90% of ad networks vet their ads to run clean, because running a malware ad is essentially a death sentence if you ever want to run any kind of premium ad (the ones that make you a lot of money) or buy premium ad space (lots of premium advertisers will specify they only want premium space, like the front page of the NYT). Above-the-board ad networks will run clean, vet their stuff, and charge a higher exchange fee, whereas unscrupulous networks (many based in eastern europe) will charge a lower fee and let all sorts of shit go through.

What does this mean? An attacker with a crafted ad that can beat cheapo mal-detection can buy cheap inventory on a shady network, intentionally outbid other people and pay a minor premium for that cheap inventory, and get their ads wherever they want. The ad network will get shut down if it was really egregious (since running a malware ad can theoretically open you to litigation from other advertisers on your network), but for every network that shuts down there's another that can pop up promising minimal overhead and minimal vetting.

The only real market solution is to whitelist a certain number of ad networks, and have sites commit to only running ads from those ad networks, but this segments the internet into the haves (premium inventory, high quality sites, premium ad networks, premium ads, all expensive) and the have nots (mom and pop sites with mediocre inventory that nobody visits because of the chance of getting cancer from the shit networks they have to run). Beyond that, this problem is unlikely to go away - it's simply too easy to game the system and put whatever you want into many adspaces.

Re:I work in online advertising

[rsborg]

Thanks for the explanation of how the advertising industry works. I really do think that commoditizing things that should really never be commoditized (i.e., home loans, ad placements, etc) creates a perverse incentive to such razor thin margins that cheating or lying becomes the only way to stay profitable.

In a larger sense, commoditization prevents competition on value. Everything competes on price, and quality isn't quantifiable as easily as price, and so there's a race to the bottom. Even if you build up a good name, a bigger player can undercut you on both price and quality for a while, drive you out of business and then completely drop the ball on quality and still rake in the profits (send a few $$ to reviewers or quality inspectors and buy a higher rating than you deserve).

Re:I work in online advertising

[RogueyWon]

The "mom and pop" sites point rings amusingly true for me.

Around a year ago, my dad went through a wave of really nasty malware infections. The ones that block your AV software, redirect your DNS and generally embed themselves right across the OS.

Now, my dad has historically been a bit of a malware-magnet. He falls into the category of "knows just about enough to think he knows everything", which used to lead him into some really poor security practices. But after a really nasty infection in 2012 which resulted in him losing quite a significant chunk of personal data, I thought he'd finally learned his lesson. He was keeping on top of Windows Update, keeping an updated AVG install, running weekly Malwarebytes scans and had finally, finally, stopped opening dodgy e-mail attachments from his perpetually-malware-infested dickhead golf-buddy friends.

I'd also put him on an adblocker. I wasn't using one myself at the time (though I am now), but I was sick of making the 4-hour-each-way journey to his place to fix his machine, so I'd held nothing back.

So a wave of four or five infections in the space of a month came as a bit of a shock. What was surprising was that he was getting re-infected very quickly after each disinfection (including one which involved a full format-reinstall of Windows).

Eventually, after going through his browser history after two consecutive infections (and half-expecting to find a megaton of pr0n), I track down the source.

And it's not pr0n, it's his bloody family history club website. Some online forum he participates in for people who are trying to trace their ancestry in a particular area. It has under 50 regular participants. It also has a prominent notice about how much the site depends on advertising income to stay in operation and asking users to disable or make an exception in their adblocker (with instructions on how to do so).

My dad has, of course, been making an exception for this site, which is then pushing a remarkably concentrated and toxic cocktail of malware-infested ads almost every time it is accessed. We actually ended up on the phone to the guy who ran the site, begging him to switch to another advertising provider. He wasn't exactly enthusiastic, so the adblocker remained in place. Don't know where things have got to since then.

Why block "in the interim"?

[gweihir]

Advertising companies obviously cannot ensure clean ads or do not care. Users are responsible for protecting their machines. The only sensible thing is to block all ads without distinction and permanently. This industry has nobody but themselves to blame for their inevitable decline.