By continuing to use the site or forum, you agree to the use of cookies, find out more by reading our GDPR policy

Every version of Windows is at risk due to a scary zero-day vulnerability after Microsoft failed to properly patch a similar flaw, a cybersecurity researcher claims. The newly discovered exploit is currently a proof-of-concept, but researchers believe ongoing small-scale testing and tweaking is setting the stage for a wider-reaching attack. “During our investigation, we looked at recent malware samples and were able to identify several [bad actors] that were already attempting to leverage the exploit,” Nic Biasini, Cisco Talos’ head of outreach, told BleepingComputer. “Since the volume is low, this is likely people working with the proof of concept code or testing for future campaigns.” The vulnerability takes advantage of a Windows Installer bug (tracked as CVE-2021-41379) that Microsoft claims to have patched earlier this month. This new variant gives users the ability to elevate local privileges to SYSTEM privileges, the highest user rights available on Windows. Once in place, malware creators can use those privileges to replace any executable file on the system with an MSI file to run code as an admin. In short, they can take over the system. Over the weekend, security researcher Abdelhamid Naceri, who discovered the initial flaw, published to Github a proof-of-concept exploit code that works despite Microsoft’s patch release. Even worse, Naceri believes this new version is even more dangerous because it bypasses the group policy included in the admin install of Windows. “This variant was discovered during the analysis of the CVE-2021-41379 patch. the bug was not fixed correctly, however, instead of dropping the bypass. I have chosen to actually drop this variant as it is more powerful than the original one,” Naceri wrote. BleepingComputer tested Naceri’s exploit and, within “a few seconds,” used it to open a command prompt with SYSTEM permissions from an account with “standard” privileges. While you shouldn’t be too worried just yet, this vulnerability could put billions of systems at risk if it’s allowed to spread. It’s worth reiterating that this exploit gives attackers admin privileges on the latest Windows OS versions, including Windows 10 and Windows 11–we’re talking about more than 1 billion systems. This isn’t a remote exploit though, so bad actors would need physical access to your device to carry out the attack. Microsoft labeled the initial vulnerability as medium-severity, but Jaeson Schultz, a technical leader for Cisco’s Talos Security Intelligence & Research Group, stressed in a blog post that the existence of functional proof-of-concept code means the clock is ticking on Microsoft releasing a patch that actually works. As it stands, there is no fix or workaround for this flaw. Naseri, who told BleepingComputer that he didn’t give Microsoft notice about the vulnerability before going public as a way to petition against smaller payouts in Microsoft’s bug bounty program, advises against third-party companies releasing their own patches because doing so could break the Windows installer. Microsoft is aware of the vulnerability but didn’t provide a timeline for when it will release a fix. “We are aware of the disclosure and will do what is necessary to keep our customers safe and protected. An attacker using the methods described must already have access and the ability to run code on a target victim’s machine,” Microsoft told BleepingComputer. To follow this thread and learn more visit OUR FORUM.

Last month, a researcher for Meta prepared a talk for colleagues that they knew would hit close to home. The subject: how to cope as a researcher when the company you work for is constantly receiving negative press. The talk had been approved to show at the company’s annual research summit for employees in early November. But shortly before the event, Meta’s legal and communications department determined that the risk of the contents leaking was too great. So it disappeared from the research summit’s agenda days before, along with another pre-taped talk describing efforts to combat hate speech and bullying. Both talks never saw the light of day. The pulling of the talks highlights how a barrage of leaks and external scrutiny has chilled the flow of information inside the company formerly known as Facebook. Many of the changes appear designed to thwart the next Frances Haugen, who worked in the Integrity organization responsible for making the social network safer before she quit earlier this year, taking thousands of internal documents with her. Those documents served as the basis for a series of damning stories in The Wall Street Journal and dozens of other news outlets, including The Verge. Some of them, such as internal research showing Instagram and Facebook can have negative effects on young people, have led to congressional hearings and lawsuits. And as the bad press continues, Meta executives have argued that the documents were cherry-picked to smear the company and paint an incomplete story. While the documents Haugen leaked haven’t yet caused Meta to make meaningful changes to its products, they’ve already left a lasting mark on how the world’s largest social network operates, particularly in its research and Integrity divisions. Ten of the 70 preapproved talks presented at the internal research summit a couple of weeks ago received a second, more stringent review to minimize leak risk. Senior leaders, including policy and communications chief Nick Clegg, have in recent months slowed the release of Integrity research internally, asking for reports to be reviewed again before they’re shared even in private groups. In some cases, researchers have been told to make clear what is defensible by data in their work and what is an opinion, and that their projects will need to be cleared by more managers before work begins. Last month, Meta rolled out a new “Integrity Umbrella” system designed to thwart leakers. The Umbrella maintains a list of employees in Integrity and gives them automatic access to join private Integrity groups in Workplace, the internal version of Facebook used by employees. When it was introduced, several employees internally pointed out that the system wouldn’t have stopped Haugen, since she worked in the Integrity division when she gathered the leaked documents. It’s not just the Integrity division that is locking down access to Workplace groups. The change has become so widespread that employees have taken to a group in Workplace titled “Examples of Meta Culture trending towards ‘Closed,’” where they’ve been posting screenshots of previously open groups they belong to being set to private. This story is based on conversations with current and former Meta employees and internal Workplace posts from the past month obtained by The Verge. In response to this story, Meta confirmed that the company was making changes to internal communication. “Since earlier this year, we have been talking about the right model of information sharing for the company, balancing openness with sharing relevant information and maintaining focus,” said Mavis Jones, a Meta spokesperson. “This is a work in progress and we are committed to an open culture for the company.” Complete details are posted on OUR FORUM.

For those of you that remember the fuss about the Y2K bug, this story may sound familiar. The Cybersecurity & Infrastructure Security Agency (CISA) has issued a warning to Critical Infrastructure (CI) owners and operators, and other users who get the time from GPS, about a GPS Daemon (GPSD) bug in GPSD versions 3.20 through 3.22. If you don’t remember the Y2K bug, let me remind you quickly. Before the year 2000, lots of computer programs kept track of the year by remembering the last two digits instead of all four. Programs coded this way would work correctly until the first day of the new millennium, when they would assume they’d been transported back in time 100 years to 1900. Some computer programs don’t care what time it is, but others do, and there were genuine fears that getting the date wrong by - 100 years might cause the lights to go out, or for planes to fall from the sky. In the end, those big problems didn’t materialize, because everyone received a warning or two, or twenty, way in advance, and there was enough time to take action and fix the broken code. Alongside telling you where in space you are, the Global Positioning System (GPS) can also tell you where in time you are. To do this, it keeps a count of the number of weeks since January 5, 1980. The main civil GPS signal broadcasts the GPS week number using a 10-bit code with a maximum value of 1,023 weeks. This means every 19.7 years, the GPS week number in the code rolls over to zero. GPSD is a GPS service daemon for Linux, OpenBSD, Mac OS X, and Windows. It collects data from GPS receivers and makes that data accessible to computers, which can query it on TCP port 2947. It can be found on Android phones, drones, robot submarines, driverless cars, manned military equipment, and all manner of other embedded systems. Unfortunately, in an echo of the Y2K bug, a flaw in some versions of GPSD could cause the time to roll back after October 23, 2021. The buggy versions of the code reportedly subtract 1024 from the week number on October 24, 2021. This would mean Network Time Protocol (NTP) servers using the broken GPSD versions would think it’s March 2002 instead of October 2021. For computer systems that have no other time reference, being thrown back in time can cause several security issues. From the perspective of incident handling and incident response, well-synchronized time across systems facilitates log analysis, forensic activities, and correlation of events. Losing track of what happened when can lead to missed incidents. Even worse is getting shut out. NTP servers using the bugged GPSD version would get thrown back almost 20 years. The Network Time Protocol (NTP) is responsible in many cases to ensure that time is accurately kept. Various businesses and organizations rely on these systems. Authentication mechanisms such as Time-based One-Time Password (TOTP) and Kerberos also rely heavily on time. As such, should there be a severe mismatch in time, users would not be able to authenticate and gain access to systems. The same would happen in cases where authentication relies on cookies. Websites and services relying on expiring cookies do not respond favorably to cookies from two decades in the future. And speaking from experience, the last GPS week number reset to zero occurred on April 6, 2019. Many GPS-enabled devices that were not properly designed to account for the rollover event exhibited problems on that date. Other equipment became faulty several months before or after that date, requiring software or firmware patches to restore their function. Since the affected versions of GPSD are versions 3.20 through 3.22 users should upgrade to version 3.23.1. Going back to older versions such as 3.19 and 3.20 is not recommended since they are unsupported and had bugs. For organizations that are using GPS appliances or rely on GPSD, it is recommended to check if GPSD is being utilized anywhere in the infrastructure and check its corresponding version. It is likely that an upgrade to GPSD will be required if no recent upgrades were performed. For more detailed information visit OUR FORUM.