Biz & IT —

TeamViewer users are being hacked in bulk, and we still don’t know how

Service blames password reuse for attacks used to drain financial accounts.

TeamViewer users are being hacked in bulk, and we still don’t know how

For more than a month, users of the remote login service TeamViewer have taken to Internet forums to report their computers have been ransacked by attackers who somehow gained access to their accounts. In many of the cases, the online burglars reportedly drained PayPal or bank accounts. No one outside of TeamViewer knows precisely how many accounts have been hacked, but there's no denying the breaches are widespread.

Over the past three days, both Reddit and Twitter have exploded with such reports, often with the unsupported claim that the intrusions are the result of a hack on TeamViewer's network. Late on Friday afternoon, an IBM security researcher became the latest to report a TeamViewer account takeover.

"In the middle of my gaming session, I lose control of my mouse and the TeamViewer window pops up in the bottom right corner of my screen," wrote Nick Bradley, a practice leader inside IBM's Threat Research Group. "As soon as I realize what is happening, I kill the application. Then it dawns on me: I have other machines running TeamViewer!"

He continued:

I run downstairs where another computer is still up and running. Lo and behold, the TeamViewer window shows up. Before I am able to kill it, the attacker opens a browser window and attempts to go to a new web page. As soon as I reach the machine, I revoke control and close the app. I immediately go to the TeamViewer website and change my password while also enabling two-factor authentication.

Lucky for me, those were the only two machines that were still powered on with TeamViewer installed. Also lucky for me is the fact that I was there when it occurred. Had I not been there to thwart the attack, who knows what would have been accomplished. Instead of discussing how I almost got hacked, I’d be talking about the serious implications of my personal data leak.

Bradley's account came a few hours after Germany-based TeamViewer reaffirmed what it has steadfastly maintained for the past two weeks—that the account takeovers are the result of end users' careless passwords practices. In a statement, company officials alluded to the recent cluster of "megabreaches" that have dumped more than 642 million passwords into the public domain over the past month. The officials wrote:

As you have probably heard, there have been unprecedented large scale data thefts on popular social media platforms and other web service providers. Unfortunately, credentials stolen in these external breaches have been used to access TeamViewer accounts, as well as other services.

We are appalled by the behaviour of cyber criminals and are disgusted by their actions towards TeamViewer users. They have taken advantage of common use of the same account information across multiple services to cause damage.

The statement went on to announce two measures being introduced in response to the large number of reported TeamViewer hijackings. The first, dubbed "Trusted Devices," ensures that before a device can access an existing TeamViewer account for the first time, the account holder must explicitly confirm that the new device is trusted. TeamViewer is implementing the measure using an in-app notification that asks account holders to approve the device by clicking a link sent through e-mail.

The second measure, called "Data Integrity," provides automated monitoring that detects when an account has been hacked.

"The system determines continuously if your TeamViewer account shows unusual behavior (e.g. access from a new location) that might suggest it has been compromised," Friday's statement explained. "To safeguard your data integrity, your TeamViewer account will be marked for an enforced password reset."

TeamViewer spokesman Axel Schmidt told Ars that TeamViewer officials initially planned to introduce these security features later this year. The growing number of public posts reporting TeamViewer account takeovers prompted the early roll out, he said.

Drinking from the firehose

Watching all the TeamViewer-related tweets and Reddit comments scroll by in real time is like drinking from a firehose. While reports of infected computers and drained accounts have reached a deafening crescendo over the past 48 hours, similar stories have been circulating for more than six months. "Teamviewer hacked to allow intruder on my desktop!" one post from December reads. "Someone got into my TeamViewer account and apparently tried to send themselves money with eBay and PayPal. What can I do to figure out what else was done?" a TeamViewer user pleaded on Reddit last month.

Many of the posts claim the takeovers are the result of a breach in TeamViewer's network, and they're being repeated so often that they're taking on the power of urban legend. A denial-of-service attack that disrupted TeamViewer's domain name system infrastructure for a few hours on Wednesday, for instance, became proof the TeamViewer domain name had been commandeered through a technique known as DNS hijacking. So far, no one has unearthed any evidence of TeamViewer's name servers using any unauthorized IP addresses, but that hasn't stopped claims like this one from circulating widely. Besides there being no factual basis for any DNS spoofing, the theory makes little sense, since the hijacking would have taken place months after the account takeovers started.

The account provided by Bradley, the IBM security researcher, is consistent with TeamViewer's position that the takeovers are the result of compromised passwords. Bradley said he had forgotten he had the remote login software installed on his computers, and the compromise was "most likely due to me not changing my leaked password."

Not that TeamViewer's public response has been much better. Representatives often go days or weeks without issuing any sort of statement, even though it's clear that a significant number of users—likely in the hundreds or thousands—are being hit by attacks that expose their most sensitive data. When company officials do respond, they issue terse press releases that omit important details. TeamViewer, for instance, has yet to address reports that some of the attacks have successfully bypassed its two-factor authentication protection, or that the attacks worked against accounts protected with strong passwords.

TeamViewer's claim that the surge in attacks is tied to the massive number of passwords that recently entered the public domain is plausible, but it's likely not the only contributing factor. It wouldn't be surprising if weaknesses in TeamViewer software are also involved. One possibility: a login mechanism that allows attackers to try large numbers of passwords without being locked out. Another: a flaw that allows attackers to circumvent two-factor protections. To date, TeamViewer's public statements leave users with a sense the company isn't providing a thorough accounting of what it knows, and that in turn gives way to mistrust and conspiracy theories.

Ars is calling on end users and network administrators who have been hit by this attack to provide log files in the hours leading up to the compromise. We'll show those files to researchers who will attempt to pinpoint common causes. Readers can submit their logs by emailing me at the the address found here.

In the meantime, TeamViewer users should ensure their accounts are protected with a randomly generated password that's at least 10 characters long, contains numbers, symbols, and upper- and lower-case letters, and is unique. It's also a good idea to run TeamViewer only when it's truly needed, rather than allowing it to autostart each time a computer is turned on. How-To Geek has a thorough guide on locking down TeamViewer here.

TeamViewer engineers certainly have the ability to perform log analyses, presumably at a much more granular level than any outsiders can. But there's more to these compromises than what TeamViewer has said to date, and it's time we all learned what it is.

Channel Ars Technica