1 00:00:01,400 --> 00:00:06,160 You can also create a. and L-M passwords using online services. 2 00:00:06,360 --> 00:00:11,800 We'll review now one of the many services of this type. 3 00:00:11,900 --> 00:00:19,650 You can enter a cryptographic resource in the form online hash Krekar dot com looking search will display 4 00:00:19,650 --> 00:00:27,070 the plaintext password used for generating the provided cryptographic resource. 5 00:00:27,110 --> 00:00:34,670 Let's see what happens if we search for one of the found hashes will choose land manager. 6 00:00:34,670 --> 00:00:38,030 This brings us back to the previous modules. 7 00:00:38,060 --> 00:00:45,740 It's not very clever to submit your own passwords to a dubious or untrusted service as you can see the 8 00:00:45,740 --> 00:00:51,900 hash has been correctly identified as a LAN Manager algorithm. 9 00:00:51,910 --> 00:00:55,700 It also turns out that the cryptographic resource was extracted from the password. 10 00:00:55,720 --> 00:00:58,450 He one. 11 00:00:58,510 --> 00:01:03,790 You can use a variety of methods to crack a password if you have an appropriate cryptographic resource. 12 00:01:07,060 --> 00:01:11,230 The showcase service also allows users to crack WPA to keys. 13 00:01:11,410 --> 00:01:14,100 Although this feature has to be paid for in most cases 14 00:01:16,810 --> 00:01:21,970 if I hash you provide is not found in the correct Tasha's database the site creators can crack it for 15 00:01:21,970 --> 00:01:25,250 you for a small fee. 16 00:01:25,250 --> 00:01:32,030 This was the last practical presentation in this module. 17 00:01:32,120 --> 00:01:35,030 What other ways can be used to obtain a user's identity. 18 00:01:36,870 --> 00:01:42,180 Replay attacks were extremely popular some time ago and even today they can still compromise Windows 19 00:01:42,180 --> 00:01:44,280 XP systems with default settings 20 00:01:46,980 --> 00:01:52,990 the attack exploits a vulnerability found in the land manager and enty land manager protocols. 21 00:01:53,130 --> 00:02:00,150 The protocols don't sign challenge in response messages and messages are not time stamped. 22 00:02:00,150 --> 00:02:02,940 This means that a message can be re-used unnoticeably 23 00:02:05,790 --> 00:02:13,570 SMB and SMB relay or programs shared on the Internet that can help you exploit this hands on an attack 24 00:02:13,570 --> 00:02:15,400 scenario is as follows. 25 00:02:19,070 --> 00:02:26,360 A client attempts to connect to a server the client has to be persuaded to connect to us. 26 00:02:26,370 --> 00:02:31,730 This can be done for example by sharing some interesting resource this attack. 27 00:02:31,730 --> 00:02:37,070 Just like the ones shown before aims to convince you that local area networks are not more secure than 28 00:02:37,070 --> 00:02:37,800 the internet. 29 00:02:39,010 --> 00:02:44,230 Sending your credentials over a local area network can be as dangerous as sending them over the Internet. 30 00:02:45,560 --> 00:02:49,270 The server should forward a challenge next since we want to connect to it. 31 00:02:51,600 --> 00:02:55,130 Instead it tells us that it wants to connect to us. 32 00:02:55,140 --> 00:02:56,680 What a coincidence. 33 00:02:56,880 --> 00:02:58,540 We react in a correct way 34 00:03:01,300 --> 00:03:04,550 following the protocol we send the challenge. 35 00:03:04,790 --> 00:03:12,200 At this point this challenge is a message that we the clients send to a server. 36 00:03:12,310 --> 00:03:16,570 Next the server realizes that we wanted to connect to it and sends a challenge message 37 00:03:19,430 --> 00:03:20,240 as it happens. 38 00:03:20,240 --> 00:03:24,320 The challenge is the same challenge that we sent to it a while before. 39 00:03:24,320 --> 00:03:25,750 We don't know that though. 40 00:03:28,540 --> 00:03:33,820 Since we have received a challenge message we encrypt it in an appropriate way using an NC or Allen 41 00:03:33,820 --> 00:03:39,160 password and for it as a response to our or the server's challenge. 42 00:03:41,010 --> 00:03:49,150 The server doesn't attempt any cryptanalysis and sends back the same response to our challenge. 43 00:03:49,260 --> 00:03:54,300 The result of this abuse is that the user pictured above in the slide has been authenticated to the 44 00:03:54,300 --> 00:04:00,510 computer of the user pictured below on the left and not the other way around. 45 00:04:00,510 --> 00:04:04,470 What's more the attackers gain the permissions of the user who started the process 46 00:04:08,130 --> 00:04:13,080 what are the countermeasures signing packets is a good solution. 47 00:04:14,230 --> 00:04:19,720 Since Allemagne and TLM don't implement this feature you need to sign packets and then they'll Aprilaire 48 00:04:20,740 --> 00:04:23,220 SMB protocol packets should be signed. 49 00:04:25,400 --> 00:04:28,660 The slide above shows how to enable SNB packet signing 50 00:04:32,120 --> 00:04:39,500 challenge response attacks have also been shown cracking the passwords was easy and fast. 51 00:04:40,820 --> 00:04:46,130 Windows has a setting that specifies which version of an authentication protocol issued by a specific 52 00:04:46,130 --> 00:04:48,020 provider will be used. 53 00:04:53,750 --> 00:05:01,780 Take a look at the options you can select sending Elham or TLM responses. 54 00:05:01,930 --> 00:05:06,430 Nothing less secure is available. 55 00:05:06,470 --> 00:05:15,950 You can also send L-M an anti alarm and if negotiated use and TLM version to if an attacker doesn't 56 00:05:15,950 --> 00:05:20,400 agree to it will send only L-M and TLM. 57 00:05:20,460 --> 00:05:24,200 The solution is just as vulnerable. 58 00:05:24,230 --> 00:05:31,510 You can also send only TLM responses without L-M this doesn't tighten security in practice however. 59 00:05:33,440 --> 00:05:38,440 The minimum authentication security level that should be set for all client computers is sending and 60 00:05:38,480 --> 00:05:42,200 TLM version 2 responses. 61 00:05:42,370 --> 00:05:48,950 If Kerberos responses fail you should only use anti-oil and version to server properties on the other 62 00:05:48,950 --> 00:05:57,000 hand should be configured to the last security level only and TLM version 2 responses should be sent 63 00:05:57,750 --> 00:06:03,610 L-M or TLM should be refused. 64 00:06:03,730 --> 00:06:09,580 If this policy is set on a server and a client has set the policy we talked about before the security 65 00:06:09,580 --> 00:06:11,850 is as high as possible without Kerberos 66 00:06:15,800 --> 00:06:18,410 Kerberos doesn't offer full protection either. 67 00:06:19,370 --> 00:06:22,430 There's no such thing as a protection guarantee with security 68 00:06:25,040 --> 00:06:31,310 Kerberos is susceptible to a certain class of attacks to attacks that compromised long term keys 69 00:06:34,020 --> 00:06:34,870 to run them. 70 00:06:34,890 --> 00:06:39,420 An attacker needs to get a hold of some cryptographic resources. 71 00:06:39,440 --> 00:06:45,230 This can be done by capturing the communications between a client that requests the TGT key and a domain 72 00:06:45,230 --> 00:06:46,880 controller. 73 00:06:46,880 --> 00:06:50,250 What happens then. 74 00:06:50,370 --> 00:06:55,290 We know already that the user's identity is verified against the message sent to and decrypted by a 75 00:06:55,290 --> 00:07:00,240 domain controller that contains a time stamp. 76 00:07:00,370 --> 00:07:07,010 The tolerance window for differences between the time stamp and the controller system time is 5 minutes. 77 00:07:07,010 --> 00:07:12,610 The problem is the time stamps always have the same structure. 78 00:07:12,710 --> 00:07:19,040 You know what the front or day month year and then hour minute second busy at the end. 79 00:07:19,300 --> 00:07:22,110 Z stands for Zulu. 80 00:07:22,260 --> 00:07:28,080 If we sniff out or obtain some encrypted resource and while trying to decrypt it using potential passwords 81 00:07:28,620 --> 00:07:33,210 we need the user passwords list again will get a value that looks like a time stamp. 82 00:07:33,210 --> 00:07:37,610 This means that the password is cracked. 83 00:07:37,670 --> 00:07:42,240 There's little chance that we'd get that time stamp value using a different key. 84 00:07:42,260 --> 00:07:50,130 This is how curb sniff works Kirsan if it's a light 10 year old application you can find on the internet 85 00:07:52,390 --> 00:07:59,240 it should be launched at the communications line between the client and I don't mean controller this 86 00:07:59,240 --> 00:08:00,920 line is relatively safe. 87 00:08:01,830 --> 00:08:08,510 Most administrators take steps to protect as you can see this protection has good grounds 88 00:08:11,380 --> 00:08:18,530 this module featured facts knowledge on varied methods of identity theft Windows system user credentials 89 00:08:18,530 --> 00:08:21,370 were used as an example. 90 00:08:21,530 --> 00:08:28,890 We showed off line attacks and outlined replay attacks online attacks. 91 00:08:29,030 --> 00:08:33,210 The module also showed the techniques and tools needed to capture a user's password. 92 00:08:33,210 --> 00:08:37,180 In about five minutes or less. 93 00:08:37,380 --> 00:08:37,900 Thank you.