101010.pl is one of the many independent Mastodon servers you can use to participate in the fediverse.
101010.pl czyli najstarszy polski serwer Mastodon. Posiadamy wpisy do 2048 znaków.

Server stats:

483
active users

#googlepasswordmanager

0 posts0 participants0 posts today

Passkeys jetzt auch auf iOS mit Chrome: Sicher und plattformübergreifend
Google hat eine wichtige Aktualisierung für seinen Google Password Manager im Chrome-Browser eingeführt. Ab sofort können Nutzer:innen von iPhone und iPad (ab iOS 17) Passkeys erstellen, ver
apfeltalk.de/magazin/news/pass
#News #ChromeIOS #FaceID #GooglePasswordManager #iCloudSchlsselbund #Passkeys #PlattformbergreifendeSynchronisation

Replied in thread

Adam, thank you for your (surprising) answer. You seem to agree with me, I'm summarizing what you wrote (quoted at the end of this toot), I joke you not:

——{
If you don't want to risk losing them, don't use ANDROID passkeys!

Instead, use a third party solution (requiring Android 14+)...
}——
 
 
*GOOGLE AUTHENTICATOR MISTAKE*
Please have a look at the weird distribution of ratings of Google Authenticator (play.google.com/store/apps/det); score : aproximate percentage of voters:

5 : 55%
sum of 2,3,4 : 20%
1 : 25% <——— note!

MOST people who voted "1", appear to have done that because, after losing (access to) their smartphone, they ALSO lost access to their (2FA TOTP-protected) accounts.

According to their reactions, most of them are PISSED; nobody warned them beforehand of this risk that TOTP secrets were not being backed up (this was changed last year; however, insecurily according to, in German, heise.de/news/Google-Authentic ).

Unfortunately, Google is making the same mistake with passkeys.
 
 
*RELIABLE LOGIN CREDS BACKUP*
Note that some security-aware people (such as I try to be) make backups of their TOTP secrets, which is POSSIBLE (I save QR-code screenshots in a password manager).

However, users CANNOT make backups of their Android passkey secrets. Therefore, if there is even the slightest chance of losing passkeys, users should ensure that a -usually PHISHABLE- alternative exists for logging in to each of their passkey-protected accounts.

Unfortunately, way too many people forget or lose "rescue codes" etc. because they hardly ever use them.
 
 
*PROMISING PASSKEY SECURITY*
The PROMISE of passkey security is relatively good, in particular for users who don't know how to choose, install (and properly configure autofill in order to prevent phishing) and use a third party password manager, and know how to backup its database (and actually make sure that this happens).

Therefore I fail to understand why it would be more important to provide an "optimal experience" to SECONDARY users of Android devices, rather than that PRIMARY users risk losing their passkeys.

Also, passwords are NOT deleted on my device when I tap "clear data"; why not?
 
 
*ARNAR WROTE*
Arnar Birgisson wrote in security.googleblog.com/2022/1 :
——{
Passkeys in the Google Password Manager are always end-to-end encrypted: When a passkey is BACKED UP, its private key is uploaded only in its encrypted form using an encryption key that is only accessible on the user's own devices. This protects passkeys against Google itself, or e.g. a malicious attacker inside Google. Without access to the private key, such an attacker cannot use the passkey to sign in to its corresponding online account.

Additionally, passkey private keys are ENCRYPTED AT REST ON THE USER'S DEVICES, with a hardware-protected encryption key.
}——
 
 
*MISLEADING DOCS/INFO*
Google's passkey documentation and your statements are incomplete, confusing and extremely inconsistent.

If passkeys are "encrypted at rest on the user's devices, with a hardware-protected encryption key", why would I care if they are synced to somebody else's account, if the other person DOES NOT POSSESS the hardware-protected encryption key?

Also, you wrote: "when they sign in on a device", "someone else signs in on their device": What Do You Mean?

Maybe someone else using the owner's screen unlock code, or signing in to an alternative Android account, or "sign in to Chrome" (whatever that means - I can imagine "signing in to" (unlocking) a /password manager) and/or switch the Google cloud account associated with the device?

As if granting another user access to your Android account on your Android device is not an extremely stupid thing to do (from a security perspective) anyway?
 
 
*JUST DON'T*
That is, unless you can trust the other user for 100% (which you never can): DON'T DO IT!

For example, your kid or grandchild may obtain access to content that your phone claims the owner is old enough for; spoofed "age verification" is just one of the increasing risks of storing "electronic passports" in smartphone "wallets". They may also steal your identity in many more ways, such as sending emails or messsges in your name, or add their credentials to your accounts (including banking apps).
 
 
*IN FACT, ARNAR AND DIANA WROTE*
Arnar Birgisson and Diana K Smetters wrote in security.googleblog.com/2023/0 :
——{
In fact, if you sign in on a device shared with others, YOU SHOULD NOT CREATE A PASSKEY THERE. When you create a passkey on a device, anyone with access to that device and the ability to unlock it, can sign in to your Google Account. While that might sound a bit alarming, most people will find it easier to control access to their devices rather than maintaining good security posture with passwords and having to be on constant lookout for phishing attempts.
}——
 
 
*CONCLUSION*
When/where did Google forget about KISS?

Why did (when Android 14 was not even available), and does Google promote passkeys - if there are even multiple ways of -unexpectedly- losing them (in my FD article I provided 3 examples) without being able to backup them by yourself?

Suppose a user, now knowing this, wants to switch from Android passkeys to, for example, Bitwarden: how do they transfer them?

Why are you not even interested in the rest of my findings?

Unbelievable.
 
 
On Feb 28, 2024, 23:30, Adam Langley (@agl) wrote:
——{
The other side of having data live on devices and using the account as a sync channel is widespread user confusion when they sign in on a device and are upset to find that their data remains on the device even after they've signed out. Or when someone else signs in on their device and their data syncs up to the other person's account.

I understand that one model isn't going to work for everybody, and Android 14 supports pluggable passkey providers so that nobody is locked into using Google Password Manager. But GPM passkeys are conceptually part of the account and clearing the account does clear them. I'll continue to try and push that our wording is consistent on this point. We'll be replacing the reset flow for passkeys in the coming months to be more specific and narrower in scope. Given that, we can be very clear about the consequences of resetting things. But while we might disagree about how Google Password Manager passkey should work, I know we do have a bug for accounts with custom passphrases. It is at least not causing data loss, but it does make the credentials inoperable. And we just need to damn well fix that and any other issues. We knew about it prior to your report but thank you for the report anyway: clear bug reports are rare.
}——

Replied in thread

@thomasbosboom : I agree that passkeys could make logging in a lot safer for a lot of people.

However, I have second thoughts w.r.t. vendor lock-in and possibly discriminating control via attestation.

Way more important, the relaliability of ANDROID passkeys is, IMO, (still) unacceptable: users risk losing their passkeys (multiple root causes), possibly resulting in being locked out of their accounts.

That is, unless a -phishable- plan B (an alternative if your passkey is no longer available) exists for each passkey-secured account. See:

"Don't use Android passkeys!" (this article I published today is in Dutch): security.nl/posting/831554/Don

Or/and my (public) answer to Adam Langley (@agl) regarding "Android passkeys unexpectedly deleted or useless after sync": infosec.exchange/@ErikvanStrat

(more details in seclists.org/fulldisclosure/20).

It may help if more people tell Adam (or his employer) that their implementation may be damaging for the acceptance of passkeys; too often people have been told "do <this> and you'll be safe".

@rmondello @denise_mochid

www.security.nlDon't use Android passkeys! - Security.NL

Overview of my noteworthy posts (as of 2024-03-07):

* My Myth series
——
Myth#3: DMARC prevents phishing
infosec.exchange/@ErikvanStrat
——
Myth#2: Ransomware myths
infosec.exchange/@ErikvanStrat
——
Myth#1: "Not my problem"
infosec.exchange/@ErikvanStrat
——
Myth#0: Authentication (factors)
infosec.exchange/@ErikvanStrat

* Android passkeys gone (latest)
——
2) My last public answer to @agl: infosec.exchange/@ErikvanStrat
——
1) Don't use Android passkeys! (rest in Dutch): security.nl/posting/831554/Don
——
0) Android passkeys unexpectedly deleted or useless after sync seclists.org/fulldisclosure/20

* Apple passkey log in without auth
infosec.exchange/@ErikvanStrat

* Awareness: check the domain name!
infosec.exchange/@ErikvanStrat

* Big Tech facilitates cybercrime
infosec.exchange/@ErikvanStrat

* TOTP does NOT improve security
infosec.exchange/@ErikvanStrat

* SMS 2FA acceptable?
infosec.exchange/@ErikvanStrat

Infosec ExchangeErik van Straten (@ErikvanStraten@infosec.exchange)Myth#3: DMARC prevents phishing A LinkedIn article [0] regarding DMARC, DKIM and SPF, written by Roger A. Grimes [1] is, IMO, an interesting read. However, I do not agree with Roger that DMARC is a succes (Note: Roger informed us about it on Mar 04, see [2]). Here's my advice: *DO* implement SPF, DKIM and DMARC - both on your sending and receiving mail servers (and consider SRS, see below). Implementing these measures (as a sender), IS important to prevent receiving mailservers (often operated by big tech) from blocking your mails. Also, it is considered BCP (Best Current Practice) to implement them on your recieving mail servers, and to honor the treatment of received mails as specified by the sender. However, w.r.t. recipients being able to distinguish between (phishing) emails SPOOFING your domain, versus emails ACTUALLY originating from your sending domain: keep your expectations at the level "FORGET IT". Details follow below, see also [3]. [0] https://www.linkedin.com/pulse/phishers-abusing-legitimate-neglected-domains-pass-dmarc-roger-grimes-wrl3e/ [1] @rogeragrimes [2] https://infosec.exchange/@rogeragrimes/112037240888788307 [3] "Forward Pass: On the Security Implications of Email Forwarding Mechanism and Policy" Authors: Enze Liu, Gautam Akiwate, Mattijs Jonker, Ariana Mirian, Grant Ho, Geoffrey M. Voelker, Stefan Savage https://arxiv.org/abs/2302.07287     ANTI-SPOOFING PROTOCOLS There are, and have been, a lot of (proposed) anti-email-abuse protocols, including SPF, micropayments [4], "Caller ID" [5][6], FairUCE [7], DomainKeys, DKIM, DMARC and ARC. [4] https://techmonitor.ai/technology/gates_backs_away_from_postage_stamps_idea_in_spam_vision [5] https://news.microsoft.com/2004/02/24/bill-gates-outlines-technology-vision-to-help-stop-spam/ [6] https://www.zdnet.com/article/anti-spam-firms-test-caller-id-for-email/ [7] https://www.crn.com/news/security/159904438/ibm-to-spam-spammers     NOT ANTISPAM Most, if not all, of these protocols were advertised as ANTI-SPAM MEASURES (by marketeers, including Bill Gates [5], who apparently failed to understand anything about them): they definitely ARE NOT THAT. In fact, many of the protocols mentioned are dead and burried, and the rest of them are, IMO, seriously ill. They are NOT antispam measures for multiple reasons, perhaps most inportantly because spammers *DO* buy their own (possibly look-a-like) domain names, and correctly set up SPF, DKIM and DMARC for those donains, and send mail from there. In fact, it may be even EASIER for spammers to set those things up than for decent domain owners, for example because the length (and nesting) of SPF records in DNS is limited. All protocols mentioned were, in fact, DESIGNED to prevent SPOOFING (forging, counterfeiting) of the sender domain (my prediction, in 2005, that SPF and FairUCE would NOT prevent spam, was spot on; see [8] - in Dutch). [8] (Dutch) https://www.security.nl/posting/10403/Waarom+SPF+en+FairUCE+op+termijn+geen+spam+bestrijden     ANTI-SPOOFING PROBLEMS Not only do the protocols mentioned not stop spam, they also, IMO unfortunately, too often, do not effectively prevent sender (domain) spoofing. Below I'll describe most of the (important, leaving out minor-) problems of SPF, DKIM and DMARC that I'm aware of.     PROBLEM #1: MUA'S AND PEOPLE An increasing number of people use MUAs (Mail User Agents, i.e. the program used by the recipient) that DO NOT show (display) the SMTP sender address (containing the sender domain) in received emails. This renders all of these measures pointless; if the recipient does not evualuate the sender domain name, ANYONE may have sent the email. Furthermore, PROVIDED THAT a MUA reveals the sender domain name, AND the recipient takes notice of it, AND the recipient knows how to interpret DNS domain names   (i.e.   accounts.google.com  differs from   accounts-google.com  (impersonator)   ), AND *magically* knows that a sender domain, such as "paypal.co.uk", ACTUALLY belongs to paypal.com - EVEN IF PayPal clearly states (on their website) that all legit mail from PayPal uses paypal.com as it's sender domain, ONLY THEN the recipient may be able to distinguish between legit and spoofed sender domains. IMO this is an EPIC FAIL, in particular because many organizations use multiple, different, sending domains (e.g. for conducting surveys). People are TOUGHT that domain names are unpredictable (i.e. that microsoftonline.com belongs to the same owners as microsoft.com, but not e.g. microsoft-sso[.]net [9]). [9]: see the IOC's at the bottom of https://www.group-ib.com/blog/0ktapus/     PROBLEM #2: SPF & FORWARDING A serious side effect of SPF is that it breaks forwarding (legitimate relaying) of emails. Note: when mentioning possible email addresses below, I use ".tld" i(nstead of the actual TLD, such as .gov or .com) For example, if a Fox News reporter "rep@foxnews.tld" mistakenly still (or wishfull thinkingly) sends an email to "trump@whitehouse.tld", and the Whitehouse's mail server forwards the mail to "trump@mar-el-lago.tld", then the Mar-el-Lago mail server's SPF implementation will typically block the mail while responding:   "The IP address of whitehouse.gov   is not permitted to send mail from   foxnews.com." The usual "SOLUTION" for this problem (called SRS = Sender Rewriting Scheme), is rather messy. Each potentially forwarding mail server, such as the whitehouse.gov's, will have to become STATEFUL. Effectively, for each to-be-forwarded email, the server must create a kind of email account (temporary and unique) with a name like "GUID_on_behalf_of_rep_at_foxnews@whitehouse.tld". This is necessary because a return path for error messages (such as mailbox full) must exist, as well as for other SMTP notifications (e.g. mail was read) generated at the Mar-el-Lago mail server. The whitehouse.gov mail server must also associate the temporary account name with the originating email address (in order for mails received by the "GUID..." account to be forwarded back to "rep@foxnews.tld"). IMPORTANT note: YOU are not in control of, potentially forwarding, receiving email servers. You may not know that they don't use SRS until you find out that your email was undeliverable.     PROBLEM #3: MAILLISTS AND DKIM Mailing list servers tend to modify emails in such a way that DKIM signature checks fail. For example, regularly "Cryptography list" emails, forwarded by the metzdowd.com mailserver, end up in my spambox with the subject line altered: "[DKIM] " is inserted at the beginning of that header line.     PROBLEM #4: DMARC: SPF *OR* DKIM The DMARC protocol is essential because, in contrast to SPF (and partially DKIM), it checks the domain name IN THE "FROM: " FIELD. That field is usually visible in MUA's,; note that SPF checks are based on the "envelope-from" domain name, which may be empty (in SMTP error messages, in order to prevent endless loops) and therefore is not fully reliable, and typically invisible in MUA's. In fact, in most (mobile) MUA's the user has no way to visualize email headers (even forwarding a probably scam-mail with headers is hard or impossible). Probably because of problems #2 and #3, DMARC is still happy if EITHER the SPF or the DKIM check (performed by the receiving mail server) fails - that is, AT LEAST ONE of the SPF and DKIM checks must succeed for the DMARC check to return a positive result.     PROBLEM #5: MS SAFE SENDERS Likely as a result of user complaints regarding not receiving mails because of DMARC failures, Microsoft USED to tell recipients to add legit senders to their list of "Safe senders". This, of course, was stupid. It meant fully bypassing SPF, DKIM and DMARC, by also delivering SPOOFED mails to the user's inbox. For example DocuSign advises customers to add docusign.com to their list of Safe Senders [10] - which used to make it easy for scammers to spoof their sender domain name. Apparenly Microsoft changed this behavior last summer: [11]. However, other recipient mail servers may still have such measures in place. [10] https://support.docusign.com/s/articles/Why-am-I-not-getting-DocuSign-email-notifications?language=en_US [11] https://techcommunity.microsoft.com/t5/exchange-team-blog/announcing-new-dmarc-policy-handling-defaults-for-enhanced-email/ba-p/3878883     PROBLEM #6: BEC Compromised sender email accounts, but also compromised "trusted third party bulk mail serviced" (example: [12]) are a big problem. [12] https://www.bleepingcomputer.com/news/security/hacked-sendgrid-accounts-used-in-phishing-attacks-to-steal-logins/     PROBLEM #7: VARIOUS OPINIONS In discussions between MTA admins regarding SPF, DKIM and DMARC, I notice significant differences in opinions which settings to use on sending MTAs (Mail Transfer Agents = mail servers) and on receiving MTA's, such as whether to use ~ or – for SPF on the sending server. Your mileage may vary... In the end, of course you are not only depending on the configuration of potentially FORWARDING mail servers, but, most importantly, on those of RECEIVING MTAs. There's not much you can do if a receiving provider (such as Microsoft) decides to (albeit conditionally) deliver mail with an SMTP sender address spoofing YOUR domain to inboxes of your recepients. In fact, admins of receiving MTAs may change settings at any time without notice (to senders AND recipients).     CONCLUSION Don't be tempted to believe that SPF, DKIM and DMARC, will prevent each recipient, in any case, to be able to distinguish between SPOOFED emails (seemingly sent by you) and emails that were REALLY sent by you. Any AUTHENTICATION check (verification of an identity) that DOES NOT effectively prevent IMPERSONATION, is hardly useful - if at all.     Edit 20240306 11:40 UTC: Changed title (old: "Myth#3: DMARC stops spoofing") and edited the first paragraph. #SPF #DKIM #DMARC #spoofing #phishing #dmarcBypass #Outlook #Microsoft #Microsoft365 #SMTP #fakeSender #forgedSender #forgedEmail #spoofedEmail #FairUCE #antiSpam #antiSpoofing #antiSpoof