Tabnabbing is a phishing method in which attackers take advantage of victims’ unattended browser tabs. After hijacking an inactive tab and redirecting it to malicious URLs, an attacker can perform a phishing attack and execute scripts.
With reverse tabnabbing, on the other hand, attackers can actually rewrite the source page after a victim clicks a malicious link. Usually, this means replacing a source page with a phishing site before the victim navigates back to that original tab. Here, the redirection happens through links from the parent site to the attacker’s site.
These attacks can fool even an attentive user. Here’s how developers can take steps to prevent them.
How does a reverse tabnabbing attack work?
A reverse tabnabbing attack begins when a victim clicks a link on a legitimate site. As the new tab opens, the attacker redirects the source page to a phishing page. After navigating back to that page, the victim thinks they are still on the right page. There, attackers can collect any credentials or sensitive data the victim enters into the decoy site.
Websites often contain external links, and if target= “_blank” is added to <a> element in HTML, the link will open in a new tab. The page linked through href might be either safe or unsafe. We do not know if the href link is a legitimate or malicious page since we have no control over it.
<a href=”page.site.com” target=”_blank”>
If a linked page is opened with target=”_blank” or by window.open() in Javascript, the linked page will have access to the same window.opener-property as the linking page. Thus, the linked page can set the property window.opener.location to any domain it wants.
Scroll to view full table
Let us check the following code snippet that contains an unsafe target blank implementation:
Vulnerable page:
<!DOCTYPE HTML>
<html>
<head>
<title>Reverse Tabnabbing</title>
</head>
<body>
<a href=”evil.sample.com” target=”_blank”>Click me</a>
</body>
</html>
Malicious site code:
<!DOCTYPE HTML>
<html>
<body>
<script>
if (window.opener) {
window.opener.location = “https://phish.sample.com”;
}
</script>
</body>
</html>
Now let’s say the victim clicks on the vulnerable page link/button. The malicious site will then open in a new tab, and the target website in the inactive/last tab will be replaced by the phishing website.
The following <a href=”https://sample.com/” target=”_blank”>link</a> is vulnerable to reverse tabnabbing because it uses target=”_blank”.
This means the page that opens in a new tab can access the initial tab and change its location using the window.opener property.
Tabnabbing in action
Let’s see this in a real-world attack scenario.
The victim has opened a social networking website called funchat.com in a browser window. They logged into the site and saw that someone had posted an offer on their wall: a great deal with a 30% discount on branded smartwatches!
The victim negligently clicks the link, which opens the eshop.com site with the offer in a new tab. However, Eshop is a malicious website that the attacker controls.
The attacker has written the code for his website like this:
<!DOCTYPE HTML>
<html>
<body>
<script>
if (window.opener) {
window.opener.location = “https://funnchat.com/login.php”;
}
</script>
<h1>AMAZING DEAL ON SMARTWATCHES!!</h1>
. . .
</body>
</html>
When the victim checks the fake offer, the malicious site forces the redirection of the victim’s original funchat.com opened tab to the fake website (funnchat.com) controlled by the attacker, which looks exactly like the funchat website’s login page.
The fake website’s login page appears and asks the victim to reenter the login credentials. Thinking they’re on the authentic site, the victim does not hesitate before entering their username and password.
The attacker has then stolen the victim’s login credentials.
Impact
This attack makes it quite probable for even a vigilant user to be lured into revealing confidential information. The user does not suspect their credentials have been stolen, simply believing they entered their password incorrectly on the fake login phishing page. The attacker can steal the victim’s login credentials and take over the victim’s account.
How to prevent reverse tabnabbing
The following fixes can help prevent this attack:
1. Add the rel=“noopener noreferrer” attribute to the links
Add rel=”noopener noreferrer” to every <a> element that has the target set to “_blank”. Noopener ensures that the linked page does not have access to window.opener from the linking page. Noreferrer makes sure that the request referrer header is not being sent. Thus, the destination site will not see the URL the user came from.
<HTML Code>
<a href=https://evil.sample.com rel=”noopener noreferrer” target=”_blank”>click here</a>
If Javascript is being used, the same can be achieved by setting the opener-property to null.
var myNewWindow = window.open(url, name, ‘noopener,noreferrer’)
myNewWindow.opener = null
If the user-generated content is being shown on the page, then sanitize the input and apply “noopener noreferrer” to every link.
2. Implement the cross-origin-opener-policy header
There is a new browser security feature called cross-origin-opener-policy (COOP). This feature can help prevent an attack where a malicious website calls “window.open” on the victim’s website and then redirects the victim to the attacker’s site.
Return the following HTTP response header from webserver. Browsers that support COOP will process-isolate the document, and attackers can’t access the victim’s site anymore:
Cross-origin-opener-policy: same-origin
3. Sandbox the frames
Sandbox the frames to prevent the tabnabbing attack from websites loaded in an iframe. Sandboxing can be achieved by setting the attribute “sandbox” as:
<iframe sandbox=”allow-scripts allow-same-origin” src=”https://www.example.com”></iframe>
The sandbox attribute controls many things by default. Mainly, it prevents the framed website from redirecting its parent site.
Take action now
Reverse tabnabbing is a severe threat with major consequences, especially when attackers carry it out as a targeted phishing attack. Luckily, you can protect your website from this kind of attack by taking the following steps:
- Add the rel= “noopener noreferrer” attribute to the links on the website.
- Implement the cross-origin opener policy header.
- Add the sandbox attribute to iframes on the website.
Nowadays, browsers support multiple security features to prevent this kind of attack. However, the developer has to take preventive measures by implementing these security controls to protect legitimate users.
Deepa Nagalingam — Security & Privacy Consultant, Penetration Testing & Ethical Hacking, IBM
Product Transformation Centre, IBM