The Future of the CAPTCHA: Interactive Advertising

I’ve recently done quite a bit of research around using CAPTCHA’s to protect against automated form submissions on web pages. Unfortunately CAPTCHA’s are a thing of the past and it’s my belief that interactive advertising is next big alternative.

CAPTCHA’s have been common place on logon and registration forms for almost a decade now. However recent research conducted by industry leaders has shown that CAPTCHA’s are no longer very effective. In fact, there are several company’s that now offer CAPTCHA solving as a service, where humans will solve them for pennies. In fact just this year Google’s Recaptcha service moved from a user input driven model to a completely data analytic model for verification. This change was largely because of one of Google’s own research fellows released a paper describing how to programmatically beat the recaptcha over 70% of the time.

So what’s next? It’s my belief that one of the new interactive advertising firm on the internet will likely soon pull through as a leader in automated form submission protection. This may sound like an unusual combination, but it really does make since. With the payouts for conventional advertising online dropping rapidly, due to increased add relestate, site owners are looking for more ways to supplement hosting costs. Advertisers are also looking for new ways to engage their audience and new social science research shows that getting people to interact with your brand makes them more likely to buy.

I guess the the other question is will interactive advertising really protect against programmatic form submissions and increase the overall security of said website? I believe if site owns force users to participate in the add and only process form submissions once they have validated with the adds API. It should still stop rapid form submissions from being processed and make it significantly harder for attackers to find injection vulnerabilities.

The only question that remains to be answered is will these interactive advertising provide enough entropy and data analytics to ensure that they can’t easily be solved or bypassed.

So You Got a Shell: SSH for Linux Post Exploitation

So you worked hard to find that cool web application exploit and got a shell on the box. Many people not so familiar with the Linux operating system, quickly find themselves throwing there hand in the air and saying; Now what do I do? By popular demand, this is a list of some of my favorite SSH for Linux Post Exploitation techniques that I’ve used over and over in past Cyber Defense competitions and my career.

Note: I’m assuming that you will likely exploit a service level account, such as ftpd, www-data, ldapadmin, squid, apache or syslog, and not immediately have root level access on the system.

One of my overall favorites is utilizing SSH for linux post exploitation,  as a mean of developing access to systems on the internal network and to maintain access. Almost all Linux distro’s come with openssh-client baked in and many also have openssh-server set to run at startup by default. Fortunately, openssh’s config files ship with fairly basic security measures in place and expect system administrators to utilize the many additional security features, that are available, to defend systems. Most simply don’t bother.  Additionally the permissions on the /etc/ssh/ssh_conf and /etc/ssh/sshd_conf files are world readable by default. So start by just checking them out.

User$ cat /etc/ssh/ssh.conf

User$ cat /etc/ssh/sshd.conf

Within /etc/ssh/sshd_conf the default configuration allows for public file authentication; with the following line.

PubkeyAuthentication yes

This line allows for a public key to be passed to the system in order to authenticate as a specified user. This is done by adding your public key to the .ssh/authorized_keys file; under the users home directory structure. Once added you can use the simply ssh to the system, using your private key, while specifying the username.

Note: Most default system service accounts ship with /bin/false or /sbin/nologin as their default, which means ssh wont allow you to login directly via ssh.

So you are looking issued the command cat /etc/passwd and she that your user has no login shell and you say Michael this whole maintaining access with public keys in the authorized_keys file is useless to me. So how about another user of SSH for linux post exploitation; using your user to port forward into the local network over ssh?

So you get a web shell using some fancy new wordpress vulnerability and then you cat out the wp-config.php file and see that the database isn’t hosted on the local system. Now you have to pop another box to access that database right? Wrong. You can create a local port forward the the mysql port with ssh. by using the following command.

ssh -R 9000:<mysql server ip>:3306 www-data@localhost

So now maybe people are thinking of great I have to have a shell again…but www-data has /sbin/nologin! Thats fine we dont need a shell. Just follow the steps above. Use your web shell from the wordpress vulnerability to use ssh-keygen to generate the www-data users keys and .ssh directly. Then simply cat your public key into the .ssh/authorized_keys file, to give you a means of local authentication. You may not be able to get a local shell, but you can still ssh port forward by telling ssh not to use interactive login or allocate a tty with the following command.

ssh -v -nNT -R 9000:<mysql server ip>:3306 www-data@localhost

This will create an ssh tunnel that will port forward all traffic from 9000 to the mysql server on its default port. So simply issue the mysql command to connect on port 9000 and use the credentials in the wp-config.php file.

mysql -u wordpress -p -h <web server ip> -p 9000

Profit!

If you have any questions or want to know more just leave a comment or hit me up on social media.

External IP Address Lookup Service

As a penetration testing having the ability to easily preform an internet facing or external IP address lookup is very important. Weather you end up on a box in a clients DMZ or you just need to set up that handler for your social engineering engagement, being able to quickly get the IP makes things a lot easier. I for one, like to do a great deal of automation and I’ve used several of these external IP address lookup services over the years. However, as I’m sure many people are aware, these services some and go like the wind.

One of my favorites that I’ve come to relay on over the years is ifconfig.me. Not only did this external ip address lookup service have some of the coolest domain names ever, but it also responded very quickly via the terminal with just the IP; while offering additionally information if you viewed the web page directly. Needless to say, I kinda just fell in love and mindlessly used its services for over a year.

Sadly ifconfig.me and several other external IP address lookup services have since slowed to a crawl or been shutdown. Likely because they begin to receive heavy traffic once they take off and generate little to no revenue for the host. Nevertheless, I’ve decided that instead of finding a new external ip address lookup service to fall in love with, I would just piggy back off my blog to release my own simple version of the site. My goal being, to maintain the external IP address lookup service, via the terminal, that I’ve come accustom to using for scripting and automation.

So without further a due I give you hackersvanguard.com/ip.php, your one stop stop for external ip address lookup. I know some people like be wondering why the heck they would ever trust me so I’ve included the complete source code bellow.

 

<?php
function getIpAddr()
{
if (!empty($_SERVER['HTTP_CLIENT_IP'])) //check to see if the ip is internal
{
$ip=$_SERVER['HTTP_CLIENT_IP'];
}
elseif (!empty($_SERVER['HTTP_X_FORWARDED_FOR'])) //check and to if the ip is being properly proxied
{
$ip=$_SERVER['HTTP_X_FORWARDED_FOR'];
}
else
{
$ip=$_SERVER['REMOTE_ADDR']; //else use the value given in the remote ip address header
}
return $ip;
}
?>