Google presented its system of using deep-learning techniques to identify malicious email attachments:
At the RSA security conference in San Francisco on Tuesday, Google’s security and anti-abuse research lead Elie Bursztein will present findings on how the new deep-learning scanner for documents is faring against the 300 billion attachments it has to process each week. It’s challenging to tell the difference between legitimate documents in all their infinite variations and those that have specifically been manipulated to conceal something dangerous. Google says that 63 percent of the malicious documents it blocks each day are different than the ones its systems flagged the day before. But this is exactly the type of pattern-recognition problem where deep learning can be helpful.
The document analyzer looks for common red flags, probes files if they have components that may have been purposefully obfuscated, and does other checks like examining macros—the tool in Microsoft Word documents that chains commands together in a series and is often used in attacks. The volume of malicious documents that attackers send out varies widely day to day. Bursztein says that since its deployment, the document scanner has been particularly good at flagging suspicious documents sent in bursts by malicious botnets or through other mass distribution methods. He was also surprised to discover how effective the scanner is at analyzing Microsoft Excel documents, a complicated file format that can be difficult to assess.
This is the sort of thing that’s pretty well optimized for machine-learning techniques.
Posted on February 28, 2020 at 11:57 AM •
In Gmail addresses, the dots don’t matter. The account “email@example.com” maps to the exact same address as “firstname.lastname@example.org” and “email@example.com”—and so on. (Note: I own none of those addresses, if they are actually valid.)
This fact can be used to commit fraud:
Recently, we observed a group of BEC actors make extensive use of Gmail dot accounts to commit a large and diverse amount of fraud. Since early 2018, this group has used this fairly simple tactic to facilitate the following fraudulent activities:
- Submit 48 credit card applications at four US-based financial institutions, resulting in the approval of at least $65,000 in fraudulent credit
- Register for 14 trial accounts with a commercial sales leads service to collect targeting data for BEC attacks
- File 13 fraudulent tax returns with an online tax filing service
- Submit 12 change of address requests with the US Postal Service
- Submit 11 fraudulent Social Security benefit applications
- Apply for unemployment benefits under nine identities in a large US state
- Submit applications for FEMA disaster assistance under three identities
In each case, the scammers created multiple accounts on each website within a short period of time, modifying the placement of periods in the email address for each account. Each of these accounts is associated with a different stolen identity, but all email from these services are received by the same Gmail account. Thus, the group is able to centralize and organize their fraudulent activity around a small set of email accounts, thereby increasing productivity and making it easier to continue their fraudulent behavior.
This isn’t a new trick. It has been previously documented as a way to trick Netflix users.
Posted on February 6, 2019 at 10:24 AM •
This vulnerability is a result of an interaction between two different ways of handling e-mail addresses. Gmail ignores dots in addresses, so firstname.lastname@example.org is the same as email@example.com is the same as firstname.lastname@example.org. (Note: I do not own any of those email addresses—if they’re even valid.) Netflix doesn’t ignore dots, so those are all unique e-mail addresses and can each be used to register an account. This difference can be exploited.
I was almost fooled into perpetually paying for Eve’s Netflix access, and only paused because I didn’t recognize the declined card. More generally, the phishing scam here is:
- Hammer the Netflix signup form until you find a gmail.com address which is “already registered”. Let’s say you find the victim jameshfisher.
- Create a Netflix account with address james.hfisher.
- Sign up for free trial with a throwaway card number.
- After Netflix applies the “active card check”, cancel the card.
- Wait for Netflix to bill the cancelled card. Then Netflix emails james.hfisher asking for a valid card.
- Hope Jim reads the email to james.hfisher, assumes it’s for his Netflix account backed by jameshfisher, then enters his card **** 1234.
- Change the email for the Netflix account to email@example.com, kicking Jim’s access to this account.
- Use Netflix free forever with Jim’s card **** 1234!
Obscure, yes? A problem, yes?
James Fisher, who wrote the post, argues that it’s Google’s fault. Ignoring dots might give people an enormous number of different email addresses, but it’s not a feature that people actually want. And as long as other sites don’t follow Google’s lead, these sorts of problems are possible.
I think the problem is more subtle. It’s an example of two systems without a security vulnerability coming together to create a security vulnerability. As we connect more systems directly to each other, we’re going to see a lot more of these. And like this Google/Netflix interaction, it’s going to be hard to figure out who to blame and who—if anyone—has the responsibility of fixing it.
Posted on April 9, 2018 at 6:30 AM •
Google has a new login service for high-risk users. It’s good, but unforgiving.
Logging in from a desktop will require a special USB key, while accessing your data from a mobile device will similarly require a Bluetooth dongle. All non-Google services and apps will be exiled from reaching into your Gmail or Google Drive. Google’s malware scanners will use a more intensive process to quarantine and analyze incoming documents. And if you forget your password, or lose your hardware login keys, you’ll have to jump through more hoops than ever to regain access, the better to foil any intruders who would abuse that process to circumvent all of Google’s other safeguards.
It’s called Advanced Protection.
Posted on October 30, 2017 at 12:23 PM •
Nice idea, but I would like it to work for other browsers and other e-mail programs.
Posted on April 1, 2015 at 1:13 PM •
Sidebar photo of Bruce Schneier by Joe MacInnis.