 Some of the most successful and lucrative online scams employ a “low-and-slow” approach — avoiding detection or interference from researchers and law enforcement agencies by stealing small bits of cash from many people over an extended period. Here’s the story of a cybercrime group that compromises up to 100,000 email inboxes daily, and apparently does little else with this access except for siphon gift card and customer loyalty program data that can be sold online. The data in this story come from a trusted source in the security industry that has visibility into a network of hacked machines that fraudsters in just about every corner of the Internet are using to anonymize their malicious Web traffic. For the past three years, the source — we’ll call him “Bill” to preserve his requested anonymity — has been watching one group of threat actors that is mass-testing millions of usernames and passwords against the world’s major email providers day. Bill said he’s not sure where the passwords are coming from, but he assumes they are tied to various databases for compromised websites that get posted to password cracking and hacking forums on a regular basis. Bill said this criminal group averages between five and ten million email authentication attempts daily and comes away with anywhere from 50,000 to 100,000 working inbox credentials. In about half the cases the credentials are being checked via “IMAP,” which is an email standard used by email software clients like Mozilla’s Thunderbird and Microsoft Outlook. With his visibility into the proxy network, Bill can see whether or not an authentication attempt succeeds based on the network response from the email provider (e.g. mail server responds “OK” = successful access). You might think that whoever is behind such a sprawling crime machine would use their access to blast out spam, or conduct targeted phishing attacks against each victim’s contacts. But based on interactions that Bill has had with several large email providers so far, this crime gang merely uses a custom, automated scripts that periodically log in and search each inbox for digital items of value that can easily be resold. And they seem particularly focused on stealing gift card data. “Sometimes they’ll log in as much as two to three times a week for months at a time,” Bill said. “These guys are looking for low-hanging fruit — basically cash in your inbox. Whether it’s related to hotel or airline rewards or just Amazon gift cards after they successfully log in to the account their scripts start pilfering inboxes looking for things that could be of value.” How do the compromised email credentials break down in terms of ISPs and email providers? There are victims on nearly all major email networks, but Bill said several large Internet service providers (ISPs) in Germany and France are heavily represented in the compromised email account data. “With some of these international email providers we’re seeing something like 25,000 to 50,000 email accounts a day get hacked,” Bill said. “I don’t know why they’re getting popped so heavily.” That may sound like a lot of hacked inboxes, but Bill said some of the bigger ISPs represented in his data have tens or hundreds of millions of customers. Measuring which ISPs and email providers have the biggest numbers of compromised customers is not so simple in many cases, nor is identifying companies with employees whose email accounts have been hacked. This kind of mapping is often more difficult than it used to be because so many organizations have now outsourced their email to cloud services like Gmail and Microsoft Office365 — where users can access their email, files, and chat records all in one place. In a December 2020 blog post about how Microsoft is moving away from passwords to more robust authentication approaches, the software giant said an average of one in every 250 corporate accounts is compromised each month. As of last year, Microsoft had nearly 240 million active users, according to this analysis. “To me, this is an important story because for years people have been like, yeah we know email isn’t very secure, but this generic statement doesn’t have any teeth to it,” Bill said. “I don’t feel like anyone has been able to call attention to the numbers that show why email is so insecure.” Bill says that in general companies have a great many more tools available for securing and analyzing employee email traffic when that access is funneled through a Web page or VPN, versus when that access happens via IMAP. “It’s just more difficult to get through the Web interface because on a website you have a plethora of advanced authentication controls at your fingertips, including things like device fingerprinting, scanning for HTTP header anomalies, and so on,” Bill said. “But what are the detection signatures you have available for detecting malicious logins via IMAP?” Microsoft declined to comment specifically on Bill’s research but said customers can block the overwhelming majority of account takeover efforts by enabling multi-factor authentication. Read the detailed report on OUR FORUM.  Microsoft will kick-start its Windows 11 rollout to mainstream users on October 5. On that day, Windows 11 will begin rolling out to eligible Windows 10 PCs and be for sale on a handful of PCs that will come preloaded with Windows 11, officials said today, August 31. There will be one feature that Microsoft originally touted as part of the Windows 11 experience that won't be available at launch: The ability to get Android apps via the Microsoft Store. Microsoft officials said today that they will have a preview of this capability ready for preview by Windows Insiders "over the coming months." Microsoft is working with Amazon and Intel on this and has been developing an Android subsystem for Windows to make it happen. Microsoft officials are planning a phased rollout of Windows 11 between October 5 and mid-2022. Microsoft will make the operating system available to new devices first. The company plans to use "intelligence models that consider hardware eligibility, reliability, metrics, age of the device and other factors" to roll it out to additional in-market PCs. Microsoft plans to use Windows Update to notify Windows 10 users when their devices are eligible to move to Windows 11. Users also will be able to "seek" manually the upgrade for eligible devices by going to Settings > Windows Update > Check for Updates. As Microsoft officials said last week, the company is testing a revamped PC Health Check app and will make it available to all users soon so they can check to see if their PCs meet Microsoft's requirements for the upgrade. Users with PCs deemed by Microsoft to be ineligible will have an option to upgrade their own PCs, with the understanding they will be in an officially "unsupported state." This means, confusingly that they may or may not get security and driver updates, according to Microsoft. Individuals or businesses who aren't ready or interested in going to Windows 11 will be able to stay on Windows 10, which Microsoft will continue to support through October 14, 2025. (There's a new Windows 10 release coming this fall, as well, which will be a very minor update known as Windows 10 21H2. Microsoft officials haven't said whether there will be a Windows 10 22H1, H2, etc.) I've asked if there will be a blocking tool, as usual, for users who don't want Windows 11 "offered" to them and their user base. Maybe one won't be needed. I was told by the company: "Microsoft is putting that choice in the hands of its customers. When a customer with an eligible Windows 10 device is notified through Window Update, they can decide if they want to upgrade to Windows 11 or stay on Windows 10." As to why Microsoft is going with October 5 as its Windows 11 release date, my guess is it's hoping to give PC makers, hoping for new Windows 11 PC purchases, a Columbus Holiday sale boost, and a long runway into the holiday 2021 season. Microsoft's blog post lists a bunch of PCs that they're touting as Windows 11-ready. None of these are unannounced devices; they're all in the market already. Follow the release of Windows 11 on OUR FORUM.  A shocking new tracking admission from Google, one that hasn’t yet made headlines, should be a serious warning to Chrome’s 2.6 billion users. If you’re one of them, this nasty new surprise should be a genuine reason to quit. Behind the slick marketing and feature updates, the reality is that Chrome is in a mess when it comes to privacy and security. It has fallen behind rivals in protecting users from tracking and data harvesting, its plan to ditch nasty third-party cookies has been awkwardly postponed, and the replacement technology it said would prevent users from being profiled and tracked turns out to have just made everything worse. “Ubiquitous surveillance... harms individuals and society,” Firefox developer Mozilla warns, and “Chrome is the only major browser that does not offer meaningful protection against cross-site tracking... and will continue to leave users unprotected.” Google readily (and ironically) admits that such ubiquitous web tracking is out of hand and has resulted in “an erosion of trust... [where] 72% of people feel that almost all of what they do online is being tracked by advertisers, technology firms or others, and 81% say the potential risks from data collection outweigh the benefits.” So, how can Google continue to openly admit that this tracking undermines user privacy, and yet enable such tracking by default on its flagship browser? The answer is simple—follow the money. Restricting tracking will materially reduce ad revenue from targeting users with sales pitches, political messages, and opinions. And right now, Google doesn’t have a Plan B—its grand idea for anonymized tracking is in disarray. “Research has shown that up to 52 companies can theoretically observe up to 91% of the average user’s web browsing history,” a senior Chrome engineer told a recent Internet Engineering Task Force call, “and 600 companies can observe at least 50%.” Google’s Privacy Sandbox is supposed to fix this, to serve the needs of advertisers seeking to target users in a more “privacy-preserving” way. But the issue is that even Google’s staggering level of control over the internet advertising ecosystem is not absolute. There is already a complex spider’s web of trackers and data brokers in place. And any new technology simply adds to that complexity and cannot exist in isolation. It’s this unhappy situation that’s behind the failure of FLoC, Google’s self-heralded attempt to deploy anonymized tracking across the web. It turns out that building a wall around only half a chicken coop is not especially effective—especially when some of the foxes are already hanging around inside. Rather than target you as an individual, FLoC assigns you to a cohort of people with similar interests and behaviors, defined by the websites you all visit. So, you’re not 55-year-old Jane Doe, sales assistant, residing at 101 Acacia Avenue. Instead, you’re presented as a member of Cohort X, from which advertisers can infer what you’ll likely do and buy from common websites the group members visit. Google would inevitably control the entire process, and advertisers would inevitably pay to play. FLoC came under immediate fire. The privacy lobby called out the risks that data brokers would simply add cohort IDs to other data collected on users—IP addresses or browser identities or any first-party web identifiers, giving them even more knowledge on individuals. There was also the risk that cohort IDs might betray sensitive information—politics, sexuality, health, finances, ... No, Google assured as it launched its controversial FLoC trial, telling me in April that “we strongly believe that FLoC is better for user privacy compared to the individual cross-site tracking that is prevalent today.” Not so, Google has suddenly now admitted, telling IETF that “today’s fingerprinting surface, even without FLoC, is easily enough to uniquely identify users,” but that “FLoC adds new fingerprinting surfaces.” Let me translate that—just as the privacy lobby had warned, FLoC makes things worse, not better. Follow this thread on OUR FORUM. |
Latest Articles
|