Targeting Lemmy specifically? probably not, but that’s not really the issue. It’s not that being a .zip address makes the server vulnerable, it’s that the existence of the .zip TLD makes everyone vulnerable:
Surveys by security researchers immediately following public release of domain registration found numerous examples of links and domains registered under .zip being used in phishing attempts, and the ICSS recommended disabling access to .zip domains until “the dust settles and risks can be assessed”.
Our findings show that the abuse rate for the .zip TLD is 0.20% which is close to the average compared to all other TLDs. This rate indicates that .zip domain names are not being used to attack users more than the average TLDs - at least for now. However, if attackers find they have better success using .zip than other TLDs, the rates of abuse might change.
Given new TLDs, such as .zip, tend to have a higher abuse rate than legacy and ccTLDs we suggest that the security research community should continue the healthy debate about the potential risks of the .zip TLD and that internet users continue to be weary of downloading and opening files with a .zip extension or TLD from sources or individuals they may not know.
The problem is that .zip conflicts with the very commonly used zip archive format which has caused user confusion - a user might click on what appears to be a URL to www.fakewebsite.zip and instead end up downloading a malicious .zip file. This creates an unnecessary and entirely avoidable security risk.
Google opened registration for the .zip and .mov top-level domains to the general public on May 3, 2023. Its release was immediately met with condemnation from cyber security experts as a result of its similarity with the file format of the same name. Malwarebytes warned against the use of already recognizable filenames and their confusion with top-level domains, as “plenty of users already have a clear idea that .zip means something completely different”. Experts cautioned against their use, and noted that the use of .zip filetypes in cybercrime had had “an explosion” in recent years. Cisco warned against the potential for leaks for personal identifying information. Researchers also registered similar concern about Google’s .mov domain.
Surveys by security researchers immediately following public release of domain registration found numerous examples of links and domains registered under .zip being used in phishing attempts, and the ICSS recommended disabling access to .zip domains until “the dust settles and risks can be assessed”.
Our findings show that the abuse rate for the .zip TLD is 0.20% which is close to the average compared to all other TLDs. This rate indicates that .zip domain names are not being used to attack users more than the average TLDs - at least for now. However, if attackers find they have better success using .zip than other TLDs, the rates of abuse might change.
Given new TLDs, such as .zip, tend to have a higher abuse rate than legacy and ccTLDs we suggest that the security research community should continue the healthy debate about the potential risks of the .zip TLD and that internet users continue to be weary of downloading and opening files with a .zip extension or TLD from sources or individuals they may not know.
Right, ok, so the problem with having a debate on this subject is that there’s no reason for this risk to exist at all. There’s no good reason to have a .zip TLD, there was no need for it, it should not have been created and no one should use it.
If you’re weighing pros and cons, there are exactly 0 pros. Therefore no matter how minor you think the cons are, they outweigh 0 pros by 100%.
Also, “nothing bad has happened yet” is not a valid argument and is a terrible basis for making risk decisions.
I see your perspective, but is there any similar instance that is not Lemmy.zip?
From another post
Lemmy.world is too big
sh.itjust.works names contains "shit", which can deter users
lemmy.ca is Canadian-centric
feddit.org, is German-centric, but technically English speaking too
dbzer0 is topic focused
programming.dev is topic-centric
blahaj is queer-focused
discuss.tchncs.de has a difficult name
lemmy.sdf.org does not defederate anyone
beehaw defederates LW and SJW
infosec.pub is topic-centric
aussie.zone is country-centric
midwest.social is region-centric
I still can’t take anyone running a .zip TLD seriously. It was bad idea to create it and it’s a bad idea to use it.
Is there any PoC of attacks on Lemmy using .zip TLD ? The instance has been up for 2 years, I never heard anything
Targeting Lemmy specifically? probably not, but that’s not really the issue. It’s not that being a .zip address makes the server vulnerable, it’s that the existence of the .zip TLD makes everyone vulnerable:
https://en.wikipedia.org/wiki/.zip_(top-level_domain)#Security_concerns
https://dnsrf.org/blog/the--zip-tld---ripe-for-abuse--but-so-far-so-good-/index.html
Not sure if that tone is the best for a healthy debate.
Can you explain why, for me? Genuinely curious, I don’t understand.
The problem is that .zip conflicts with the very commonly used zip archive format which has caused user confusion - a user might click on what appears to be a URL to www.fakewebsite.zip and instead end up downloading a malicious .zip file. This creates an unnecessary and entirely avoidable security risk.
https://en.wikipedia.org/wiki/.zip_(top-level_domain)#Security_concerns
Choosing to use this TLD basically just screams ignorance, and should be causing users to question the competence of the person who made that choice.
https://dnsrf.org/blog/the--zip-tld---ripe-for-abuse--but-so-far-so-good-/index.html
Not sure if that tone is the best for a healthy debate.
Right, ok, so the problem with having a debate on this subject is that there’s no reason for this risk to exist at all. There’s no good reason to have a .zip TLD, there was no need for it, it should not have been created and no one should use it.
If you’re weighing pros and cons, there are exactly 0 pros. Therefore no matter how minor you think the cons are, they outweigh 0 pros by 100%.
Also, “nothing bad has happened yet” is not a valid argument and is a terrible basis for making risk decisions.
I see your perspective, but is there any similar instance that is not Lemmy.zip?
From another post
https://lemmy.dbzer0.com/post/37336391?scrollToComments=true
https://lemmyverse.net/?order=active_month
‘This bridge is literally held together with duct tape, but it hasn’t killed anyone yet!’
I’m with you, unecessary risk. Thank you for the explanations.