This article is more than 1 year old
Time for people to patch backup plugin for WordPress
Plus advice for Cisco admins from the NSA and blurring's not the best
In brief If you're using the UpdraftPlus WordPress plugin to back up your systems, you'll need it patched – or else risk sharing your backups with strangers.
The UK-based plugin producer warned customers on Thursday to upgrade to version 1.22.3 of the code after Marc Montpas, a security research engineer at development house Automattic, spotted a potentially serious mistake. After disclosing it responsibly, the biz had a fix ready in two days.
"This defect allows any logged-in user on a WordPress installation with UpdraftPlus active to exercise the privilege of downloading an existing backup, a privilege which should have been restricted to administrative users only," UpdraftPlus' advisory states.
"This was possible because of a missing permissions check on code related to checking current backup status. This allowed the obtaining of an internal identifier which was otherwise unknown, and could then be used to pass a check upon permission to download."
Thankfully, the attack vector is only open to any logged-in user, and its relative complexity (Montpas' full breakdown is worth a read,) mean this wouldn't be used for a large-scale attack, despite the millions of users. Rather it's more likely to be used very selectively, at least on those that haven't patched.
Mexican corporate snooper pleads guilty to US spyware kit shenanigans
Carlos Guerrero, 48, has pleaded guilty to a single charge of conspiracy, admitting he actively helped sell hardware and software used to spy on Americans.
Guerrero admitted to taking around $25,000 from a "large Mexican business" to provide access to the emails and phone calls of a Florida-based sales staffer. He also said he used his own products to spy on a US rival to his own business.
He was active south of the border too, selling tools to a Mexican government official, knowing they would be used to spy on the politician's political rival's Twitter, Hotmail, and iCloud accounts. He also sold to numerous other private individuals and faces a possible five years in prison.
The operation was busted as part of a cross-border investigation into the sale and resale of hacking tools by the Homeland Security Investigations unit. Guerrero admitted to setting up a reseller deal with an unnamed Italian company in 2014 that sold spyware and surveillance hardware, and then doing similar business with an unnamed Israeli company – make your own guesses on the names.
"With this guilty plea, we are sending a clear message that companies and individuals who unlawfully violate privacy rights will not be tolerated and they will be held accountable," said Chad Plantz, special agent in charge for HSI San Diego.
"The world we live in is increasingly interconnected by technology meant to improve our lives, but as seen in this case, this same technology can be acquired by bad actors with harmful intentions."
Blurring just won't cut it
Using a blurring function to hide information is common enough, but increasingly insecure, and now you can test documents yourself.
But better technology and smarter software suggests a simple blur function is woefully inadequate. Last year security consultants Jumpsec issued a challenge text and invited people to deblur the message, and this week Dan Petro, a lead researcher at infosec shop Bishop Fox, confirmed he'd broken the blurring.
Petro isn't releasing the full text of the blur, so as not to spoil the surprise. He's also released an open-source version of the code he used, dubbed Unredactor, so people can test out what can be found and provide feedback. The full breakdown of how it's done has now been published (watch out as the video does contain swearing) and the end message is very clear.
"The bottom line is that when you need to redact text, use black bars covering the whole text. Never use anything else," he recommended. "No pixelization, no blurring, no fuzzing, no swirling. Oh, and be sure to actually edit the text as an image.'
NSA has a problem with Cisco passwords
The US National Security Agency has issued a briefing document specifically for Cisco users explaining how to make their passwords stronger.
The advisory [PDF] recommends only one type of password, Cisco's Type 8, which uses either Password-Based Key Derivation Function version 2 (PBKDF2), SHA-256, an 80-bit salt – one NSA wit described it as "what Type 4 was meant to be," in the document.
For second best there's Type 6, which uses an 128-bit AES algorithm and is particularly useful for VPN passwords, the NSA notes, although Type 8 is preferable.
- US to attack cyber criminals first, ask questions later – if it protects victims
- HPE, Samsung invest in quantum startup Classiq
- Linux Snap package tool fixes make-me-root bugs
- Adobe warns of second critical security hole in Adobe Commerce, Magento
"Type 8 should be enabled and used for all Cisco devices running software developed after 2013. Devices running software from before 2013 should be upgraded immediately," the agency warned. "Type 6 passwords should only be used if specific keys need to be encrypted and not hashed, or when Type 8 is not available (which typically implies that Type 9 is also unavailable)."
On the absolute "Do not use" list are Type 0 (plain text), Type 4 – which uses a crippled form of PBKDF2 that's susceptible to brute forcing – and Type 7, a Vigenere cipher that can be easily broken.
As for other advice, the NSA recommends choosing long-form passwords and admins limiting access privileges for users more strictly, but you know that already.
Teams' week gets worse
Just days after leaving some Teams users unable to make and receive calls, thanks to a dodgy update, a new report from cloud security shop Avanan is warning that Microsoft's platform has been spreading malware.
The campaign was spotted in January and is ongoing, Avanan said, and uses phished credentials to access Teams chat sessions. Once there, the attackers will drop innocuously named .exe files onto forums in the hope someone will click on them, thus installing DLL files on their machine that can be used to set up remote access by the miscreants.
"Once inside an organization, an attacker usually knows what technology is being used to protect it," the report said. "That means they will know what malware will bypass existing protections. Compounding this problem is the fact that default Teams protections are lacking, as scanning for malicious links and files is limited."
"We offer a default layer of protection that includes malware scanning for shared files and we encourage all customers to investigate and implement additional layers of protection and apply best practices depending on their unique needs," a Microsoft spokesperson told The Register. ®