That holiday deal in your social media feed may not be the bargain it appears to be. Scammers are taking advantage of the busy shopping season, and new technology can make fake messages harder to spot.
The big red flag is anything that catches you off guard. An unexpected text, email or social media ad should make you stop before clicking, especially if it's pushing a deal that seems too good to be true.
WATCH: AI makes holiday shopping scams harder to spot, expert warns
AI Makes Holiday Shopping Scams Harder To Spot Here's What To Watch ForBrian Smith with Redstone Federal Credit Union says shoppers should stick with retailers they know. Instead of following a link from a text or social media post, go directly to that retailer's legitimate website and see if the same offer is actually there.
If you think you've fallen for a scam, contact your bank or credit union immediately. You do not have to wait for a suspicious charge to clear before asking for help. And don't feel bad if you fall for something the technology scammers are using is different this year.
"With the advancements of AI, the impersonation fraud- and what I mean by that is AI has the ability to impersonate individuals, it has the ability to impersonate companies, and the text messages and the communications can look very valid," Smith said.
That means some of the old warning signs, like obvious spelling or grammar mistakes, may not be there anymore.
Before you start shopping, check your banking app. Smith recommends turning on transaction alerts and using card controls if your bank offers them, so you can spot suspicious activity more quickly.
Have you spotted a suspicious "deal" in your feed this season? See the exact warning signs Brian Smith says to watch for and what to do if you've already clicked in the full video report above. Then share the scam that almost got you by emailing me directly at Eric.Pointer@NewsChannel5.com, and your tip could help protect your neighbors this holiday season.
This story was reported by Eric Pointer and has been converted to this platform with the assistance of AI. Eric Pointer and our editorial team verifies all reporting on all platforms for fairness and accuracy.A newly disclosed fake U.S. construction and engineering firm is attributed to a North Korean remote worker. The operation used a corporate website and an apparent real-time face swap to pursue contracts.
The firm, called The Great House, presented itself as a residential and commercial contractor based in Lynchburg, Virginia. In its November 19, 2026, report, Silent Push assessed that the operation sought to generate revenue for North Korea.
Silent Push examined one front company and held two calls with its operator; they reported no awarded contracts or payments.
Analysts traced thegreathouse[.]us through a WHOIS registrant pivot, using an email address from a publicly reported list of roughly 1,400 addresses attributed to North Korean IT workers.
The search returned one domain, registered on October 18, 2025, using devtop0109[@]gmail[.]com. Researchers also reviewed historical web scans and the company’s LinkedIn presence.

The site claimed more than a decade in business, over 200 clients, and more than 60,000 square feet built. The registration date did not disprove those claims, but it gave procurement teams a reason to verify the portfolio independently.
The contact page also exposed a discrepancy between rendered text and HTML markup. It displayed a Virginia number with a 434 area code, but the tel: hyperlink dialed an Arkansas number with a 479 area code.
The registration record listed a Massachusetts number with a 351 area code and a New York address. Posing as prospective clients, analysts contacted the Telegram account listed on the site.
The operator, who introduced himself as “Edward,” kept his camera off throughout the first call, which lasted about 25 minutes. Edward resisted identity verification but agreed to appear on camera for the second meeting.
Analysts observed compositing artifacts consistent with a live facial overlay, including visible layering and distortions they described as facial “folding.” He repeatedly kept his mouth out of frame. When he leaned back, the inconsistencies became easier to see.
He declined to pass a hand in front of his face, an occlusion test that can expose overlay errors, and soon cut the video feed. The technique resembles real-time deepfakes used in remote job interviews.
Silent Push compared the corporate facade with Contagious Interview’s fake consulting firms, which used fabricated staff profiles and malicious repositories to deliver BeaverTail, InvisibleFerret, and OtterCookie to job seekers.
The construction front reversed the approach: rather than recruiting developers, it presented itself as a vendor bidding for paid work. Researchers observed no malware delivery during the engagement.
The case also differs from North Korean laptop-farm operations, in which overseas workers remotely controlled employer-issued devices hosted at U.S. addresses.
Silent Push recommends checking contractor licenses directly with state licensing boards and reconciling company addresses, phone numbers, and domain registration records before onboarding.
During video calls, reviewers can ask representatives to turn sideways or briefly obscure their faces, then watch for distorted edges or audio and lip movements that fall out of sync.
A refusal warrants further review, not an attribution. U.S. organizations can report suspected North Korean remote-worker activity to the FBI’s Internet Crime Complaint Center (IC3).
Join 16,000+ SOC teams using ANY.RUN to streamline threat investigations and reduce manual effort. Explore for your team
The post North Korean Hackers Use Deepfake Video Calls and Fake US Construction Company to Secure Contracts appeared first on Cyber Security News.
An improper authentication flaw in the WatchGuard Dimension log collection service allows remote code execution without valid credentials.
Tracked as CVE-2026-86130, the flaw carries a reported score of 8.7 and affects Dimension versions 2.0 through releases earlier than 2.3.2. The advisory details identify version 2.3.2 as the first fixed release for all 15 listed CVEs.
Three other findings carry higher reported scores of 9.3, but the log collector flaw stands out because it does not require an existing account or administrative session.
CVE-2026-86130 affects Dimension’s log ingestion service, not the Fireware VPN subsystem implicated in the WatchGuard Firebox IKEv2 code execution flaw.
The immediate exposure question is which networks can reach the collector, even when the web console itself is restricted. A separate collector finding, CVE-2026-93911, also carries a reported score of 8.7 but stems from SQL injection rather than improper authentication.
The advisory summaries do not establish that the two findings share an entry point, require the same privileges, or execute code under the same account.
Two session hijacking findings expose tokens through different components: CVE-2026-84322 affects the management API, while CVE-2026-86094 affects diagnostic logs. CVE-2026-81434 introduces reflected cross-site scripting on the device configuration history page.
CVE-2026-86126, a cross-site request forgery flaw in user management, allows creation of a Super Administrator account.
Those findings affect session confidentiality and administrative authorization, distinct from the service-level authentication failure behind the WatchGuard Agent execution flaws. CVE-2026-93916 separately allows Super Administrator access to persist because sessions do not expire as required.
The remaining execution paths span log notification processing, a system settings endpoint, email server configuration, database snapshot restoration, and package updates.
The supplied descriptions explicitly identify root-level execution for CVE-2026-93914 and CVE-2026-103580, but do not establish root privileges for the unauthenticated collector flaw.
Other findings affect system integrity or confidentiality without claiming code execution: CVE-2026-81430 permits arbitrary file writes during system upgrades, CVE-2026-93917 permits takeover of managed devices through improper registration authorization, and CVE-2026-126037 exposes device summary data through broken access control.
Administrators should inventory Dimension instances, prioritize collectors reachable from untrusted networks, and upgrade affected deployments to 2.3.2 or later.
WatchGuard stated that until the upgrade is complete, limit collector access to networks that send logs and restrict the management interface to authorized administrative networks.
As with the separate WatchGuard access point management API flaws, network isolation reduces reachability but does not correct the underlying defects. After deployment, verify the installed version and confirm that expected log sources are still reporting.
For suspected compromises, preserve diagnostic logs, administrative audit records, and relevant network telemetry before making changes that could overwrite evidence.
Review unexpected Super Administrator creation, unexplained device registration changes, and abnormal activity involving snapshots, packages, or notifications.
Investigate possible token exposure separately from password compromise: check whether affected sessions remain active and invalidate them where possible.
| CVE | Technical finding | Reported impact | Reported score |
|---|---|---|---|
| CVE-2026-81430 | Arbitrary file write in system upgrade process | Unauthorized file writes | 4.8 |
| CVE-2026-81434 | Reflected cross-site scripting on device configuration history page | Script execution in the browser | 9.3 |
| CVE-2026-84322 | Session token exposure in management API | Session hijacking | 9.3 |
| CVE-2026-86094 | Session token exposure in diagnostic logs | Session hijacking | 8.6 |
| CVE-2026-86126 | Cross-site request forgery in user management | Super Administrator account creation | 9.3 |
| CVE-2026-86130 | Improper authentication in log collection service | Unauthenticated remote code execution | 8.7 |
| CVE-2026-93911 | SQL injection in log collector | Remote code execution | 8.7 |
| CVE-2026-93912 | Code injection in log notification processing | Remote code execution | 8.6 |
| CVE-2026-93914 | Improper input validation in database snapshot restoration | Code execution as root | 7.7 |
| CVE-2026-93915 | Missing authorization in system settings endpoint | Remote code execution | 8.6 |
| CVE-2026-93916 | Insufficient session expiration | Persistent Super Administrator access | 8.6 |
| CVE-2026-93917 | Improper authorization in device registration | Managed device takeover | 8.6 |
| CVE-2026-103579 | Carriage return and line feed (CRLF) injection in email server configuration | Remote code execution | 8.6 |
| CVE-2026-103580 | OS command injection in package update | Remote code execution as root | 5.3 |
| CVE-2026-126037 | Broken access control in device summary | Unauthorized data disclosure | Not provided |
Join 16,000+ SOC teams using ANY.RUN to streamline threat investigations and reduce manual effort. Explore for your team
The post Critical WatchGuard Dimension Flaw Lets Hackers Execute Code Without Authentication appeared first on Cyber Security News.