Moonpig breach highlights need for app and API testing

A severe vulnerability in the API used by Moonpig's Android app has highlighted the need for organisations to apply greater scrutiny to the security of their apps and endpoints. Through its apps and website, the custom greetings card company sends out more than 12 million cards every year and turned over £53 million last year.

By enumerating an easily predictable sequence of user ID numbers, anyone could retrieve various information about millions of Moonpig customers, including names, addresses, and some credit card details. Because there was no authentication mechanism for the API, an attacker could also have placed orders on other customers' accounts.

Unlike with traditional web applications, much of what goes on beneath the glossy facade of an app is hidden from the user — but with the right tools and the right knowledge, it can be trivial to identify and exploit any vulnerabilities that might affect it. The Moonpig vulnerability exemplifies this, as the problem was not only easy to spot, but could be exploited simply by pasting a modified URL into a standard web browser.

The Moonpig vulnerability stemmed from the fact that the API trusted data sent from the app, without considering that it could have been altered or fabricated by a malicious party. This type of vulnerability fundamentally compromises the security of the application and the data it handles, and would likely be quickly identified in a third-party security test of the API.

The danger posed by this vulnerability was compounded by Moonpig's failure to react promptly — Moonpig purportedly knew about this issue 17 months ago after it was reported by one of its own customers. However, Moonpig failed to shut down or fix the vulnerable service until after the vulnerability was publicly disclosed last night.

Moonpig issued the following statement on its website today:

You may have seen reports this morning about our Apps and the security of customer details when shopping with Moonpig. We can assure our customers that all password and payment information is and has always been safe. The security of your shopping experience at Moonpig is extremely important to us and we are investigating the detail behind today's report as a priority. As a precaution, our Apps will be unavailable for a time whilst we conduct these investigations and we will work to resume a normal service as soon as possible. The desktop and mobile websites are unaffected.

Netcraft offers Mobile App Security Testing services and traditional Web Application Security Testing, both of which include testing of relevant APIs and other endpoints that may be commonly overlooked. Contact us at to discuss your requirements.

Most Reliable Hosting Company Sites in December 2014

Rank Performance Graph OS Outage
DNS Connect First
1 Qube Managed Services Linux 0:00:00 0.007 0.097 0.037 0.075 0.075
2 Datapipe Linux 0:00:00 0.011 0.104 0.017 0.034 0.047
3 Anexia Linux 0:00:00 0.011 0.430 0.088 0.195 0.216
4 CWCS Linux 0:00:00 0.011 0.226 0.099 0.181 0.183
5 iWeb Linux 0:00:00 0.019 0.132 0.079 0.156 0.156
6 Linux 0:00:00 0.022 0.150 0.073 0.157 0.203
7 ServerStack Linux 0:00:00 0.022 0.087 0.079 0.156 0.156
8 LeaseWeb Linux 0:00:00 0.026 0.170 0.026 0.058 0.058
9 Bigstep Linux 0:00:00 0.026 0.126 0.060 0.122 0.122
10 Linux 0:00:00 0.026 0.149 0.062 0.141 0.141

See full table

Qube Managed Services had the most reliable hosting company website in December, with just two failed requests. London-based Qube has been in the top 10 for all but one month in 2014, and reached first place six times. Qube offers a range of managed services, including colocation, cloud hosting and dedicated servers. It has data centres in London, New York and Zurich and focuses on providing reliable and high-quality solutions to its clients.

Datapipe follows closely in second place, with three failed requests. Datapipe, whose website has over 8 years of continuous uptime, was recently named a leader in hosted private cloud solutions by a Forrester Research report. Stratosphere, Datapipe's hosted private cloud solution, supports multiple hypervisors and has data centres in seven locations spread throughout the world.

Anexia, which also had three failed requests but a higher average connection time, came in third place this month. Anexia last entered the top ten in July 2014 when it took first place. It was recently named by Deloitte as one of the fastest-growing Austrian technology companies of 2014, marking the second year in a row that it has appeared in Deloitte's Technology Fast 500 EMEA list.

CWCS closely follows Anexia in fourth place, having the same number of failed requests but a slightly higher average connection time. UK-based CWCS last featured in the top ten in August 2013.

Linux remains the most popular choice of operating system, with every hosting company website in December's top ten served by Linux machines. Notably, Datapipe, whose website previously used FreeBSD, switched to Linux in November.

Netcraft measures and makes available the response times of around forty leading hosting providers' sites. The performance measurements are made at fifteen minute intervals from separate points around the internet, and averages are calculated over the immediately preceding 24 hour period.

From a customer's point of view, the percentage of failed requests is more pertinent than outages on hosting companies' own sites, as this gives a pointer to reliability of routing, and this is why we choose to rank our table by fewest failed requests, rather than shortest periods of outage. In the event the number of failed requests are equal then sites are ranked by average connection times.

Information on the measurement process and current measurements is available.

Proxy auto-config attacks defeat 2-factor auth, hide using country specific content

Fraudsters have been using proxy auto-config (PAC) scripts to steal online banking credentials for several years, but as with most phishing techniques, it is inevitable for these attacks to evolve and become more effective. The latest spate of PAC attacks has achieved this by using geolocation technology to evade detection and select which targets to attack.


PAC attacks typically channel online banking traffic through rogue proxy servers, allowing fraudsters to gobble up unencrypted usernames and passwords when forms are submitted, or to hijack already-authenticated sessions by stealing session cookies. Being able to view and modify this traffic also allows two-factor authentication mechanisms such as one-time passwords to be easily defeated.

PAC scripts used in these attacks inevitably look suspicious, which highlights the fact that fraud is taking place. Consequently, it is in the fraudster's interest to stop these scripts being found by law enforcement agencies, or indeed anyone else who might be tasked with investigating or preventing the fraud.

The latest attacks use a PAC script which is hosted on a web server in the Netherlands. This server has been configured to refuse TCP connections from certain countries or locations, which could be sufficient to put an investigator off the scent – if the server simply does not appear to exist, they may not bother investigating further. Meanwhile, the remaining unblocked users will continue to fall victim to the PAC attack.

Where the server can be accessed, geolocation is also used to customise the contents of the PAC script. For example, a completely benign PAC script is returned to clients in Australia, which simply tells the victim's browser to connect directly to all websites; no proxying takes place:

Deobfuscated JavaScript from the benign PAC script

Deobfuscated JavaScript from the benign PAC script

Conversely, requesting the PAC script from Japan causes the following JavaScript to be returned:

PAC attack against Japanese banking customers (contents  deobfuscated for clarity)

PAC attack against Japanese banking customers (contents deobfuscated for clarity)

The FindProxyForURL function specifies which hostnames should be proxied through the fraudster's server. Anyone using this proxy script will be giving the fraudster an opportunity to observe or modify all unencrypted traffic flowing between his browser and each of the specified Japanese online banking websites.

If the victim browses to a site which does not match any of these patterns, his browser will not use the proxy and instead make a direct connection to be site. This serves to reduce the load on the fraudster's proxy server, as well as reducing the likelihood of the victim noticing something is awry. For example, if the victim performs a Google search for "what is my ip?", his browser will connect directly to, causing Google to display the victim's own IP address rather than that of the fraudster's proxy.

Although online banking sites are the clear targets of these attacks, it is notable that many of these scripts, including the Japanese example, also target Facebook. The following PAC script is returned to clients in Switzerland, and proxies traffic destined for *, as well as several Swiss banking websites.


It was not apparent why Facebook is being targeted among these banks, but compromised Facebook accounts could be useful for propagating the malicious proxy scripts to other users. For example, users could be tricked into manually editing their proxy settings by following instructions posted from a trusted friend's compromised account, or other social engineering tricks to get the user to download and run malware.

This PAC attack is still active, with Japan and Switzerland being targeted by distinct malicious scripts. Most locations are unable to connect to the Dutch PAC script server, apart from Australia and Poland, which receive an identical benign script which does not proxy any web traffic.


Poor web application security can contribute significantly to the success of these proxy-based attacks. For instance, if the session cookies used on a bank's HTTPS website are not marked with the Secure attribute, then they will be transmitted unencrypted through the fraudster's proxy if the victim subsequently makes an HTTP request to the same hostname. Such attacks are much less likely to succeed if the targeted HTTPS site uses HTTP Strict Transport Security (HSTS) to prevent the connection being downgraded to HTTP.

Netcraft's Web Application Security Testing service can identify sites that are readily vulnerable to these types of attack. Banks and other organisations can also use Netcraft's takedown service to remove malicious proxy scripts and phishing sites from the internet, while infrastructure providers can use our phishing site feed to protect their users. For more information, please contact

December 2014 Web Server Survey

In the December 2014 survey we received responses from 915,780,262 sites and 5,034,578 web-facing computers.

This is the second month in a row where there has been a large drop in the total number of websites, giving this month the lowest count since January. As was the case in November, the loss has been concentrated at just a small number of hosting companies, with the ten largest drops accounting for over 52 million hostnames. The active sites and web facing computers metrics were not affected by the loss, with the sites involved being mostly advertising linkfarms, having very little unique content. The majority of these sites were running on Microsoft IIS, causing it to overtake Apache in the July 2014 survey. However the recent losses have resulted in its market share dropping to 29.8%, leaving it now over 10 percentage points behind Apache.

Despite losing more than six million hostnames this month, nginx outpaced all other major server vendors by gaining 22,300 web-facing computers. nginx is now used by nearly 11% of all web-facing computers – twice the share that it had two years ago.

Overall, the total number of web-facing computers in our survey increased by just over 40,000 this month, making nginx responsible for more than half the increase. Despite an increase of over 11,000 computers for Apache, and 1,700 for IIS, both continue to lose market share.

Thanks to continued strong growth at Amazon Web Services, Amazon is the largest hosting company by a considerable margin in terms of our web-facing computers metric (which includes web-facing virtual machines, providing that each has its own kernel and IP address). With nearly 300,000 web-facing computers in total, Amazon has just over twice as many as second-place OVH. In October, we reported that DigitalOcean had become the 4th largest hosting company in under 2 years, but it quickly reached third place in November and is continuing to close the gap on OVH.

Cloud growth

Both Amazon Web Services and Microsoft Azure expanded their cloud hosting footprints recently. Amazon opened a new European AWS region in Frankfurt, which augments its existing EU region in Ireland. Besides being able to host services closer to the center of Europe, the new region means that customers can now build multi-region applications with the assurance that their data will stay within the EU. The new Frankfurt region houses two EC2 availability zones and three AWS edge locations.

Microsoft's new Azure "geo" is in Australia, and consists of two geographically redundant regions in New South Wales and Victoria. This will help Microsoft to compete with Amazon's two EC2 availability zones in Sydney.

New TLDs

More new top-level domains showing strong growth in this month's survey include .nyc, which is targeted for use by New Yorkers, and .realtor, which is only allowed to be used by members of the National Association of Realtors or the Canadian Real Estate Association. These have grown from virtually nothing to a total of 40,000 and 80,000 sites respectively.

Ironically, one of this month's fastest growing new top-level domains started off as an April Fool's Day joke in 2009, when the founder of OVH announced the creation of the .ovh TLD – years before such things were actually possible. This joke resulted in over 22,000 requests to register .ovh domains within a few hours, demonstrating the potential demand for such domains. OVH eventually entered into a Registry Agreement with ICANN in January 2014, and the sunrise period for the .ovh TLD began in September. This month's survey saw the number of sites using .ovh domains grow from 6,000 to 63,000, likely due to the first 50,000 .ovh domains being given away for free, and with subsequent growth being fuelled by attractive pricing: new .ovh domains can currently be registered for only EUR 0.99 per year and renewed for EUR 1.99.

Total number of websites

Web server market share

DeveloperNovember 2014PercentDecember 2014PercentChange
Continue reading

ICANN hit by successful spear phishing attack

The Internet Corporation for Assigned Names and Numbers (ICANN) has fallen victim to a phishing attack which resulted in the attackers gaining administrative access to some of ICANN's systems, including its Centralized Zone Data Service (CZDS).

In an email alert sent this morning, ICANN said it believes a spear phishing attack in November resulted in several ICANN staff members' email credentials being compromised. The stolen passwords were then used to gain unauthorised access to multiple ICANN systems, which could have resulted in other usernames and passwords being compromised.

Although CZDS passwords are stored as salted hashes, ICANN has taken the precaution of deactivating passwords and API keys used on the compromised CZDS service. ICANN implemented some security enhancements earlier this year, which it believes limited the extent of the unauthorised access, and has implemented further measures since this attack.

The spear phishing emails involved in this attack were crafted to appear to originate from ICANN's own domain, which is a common tactic for phishers as it lends a fair amount of credibility to the emails. This domain spoofing could well have played an important part in the successfulness of the attack, but still does not feature any Sender Policy Framework records to specify who can send mail on its behalf.

Organisations concerned about these types of attack can use Netcraft's Fraud Detection service, which processes DMARC (Domain-based Message Authentication, Reporting and Conformance) reports on your behalf. These reports are sent by ISPs and e-mail receivers when they see any emails which claim to be from one of your own domains. A web interface shows the status of all of your own domains, any configuration changes required, and highlights unprotected domains being used by fraudsters attacking your customers.

Banks allow phishers to log in using Tor

The Financial Crimes Enforcement Network (FinCEN), a department of the US Treasury that combats financial crimes such as fraud and money laundering, recently released a report stating that "nearly $24 million in likely fraudulent activity" involved known Tor network nodes. The proportion of fraud that involves Tor is increasing rapidly: according to the report, October 2007 to March 2013 saw an increase of 50% in Tor-related fraud reports, whereas the most recent and much shorter period of March 2013 to July 2014 saw an increase of 100%. The report, which is not public, was obtained by computer security journalist Brian Krebs.

Tor is a piece of open-source software that attempts to provide online anonymity using a technique known as "onion routing". Messages sent by the user, such as HTTP requests from the user's web browser, are sent across the Tor network, instead of being sent directly to the destination server. Before a user sends a message, it is encrypted several times, along with information describing how the message should be routed through a virtual circuit across the Tor network. Circuits consist of a series of three randomly-selected Tor nodes: an entry node, a middle node and an exit node. The user's traffic enters the Tor network at the entry node. Each successive node is able to remove a single layer of encryption, which also reveals the next node to send the message to – akin to peeling the layers of an onion. When the message reaches the exit node, the final layer of encryption is removed and it is sent out across the Internet to its final destination. A similar procedure applies to messages travelling in the opposite direction back to the user, such as HTTP responses.

A diagram showing the nodes and the links between them in a Tor circuit. Although Tor does not encrypt the communication between the exit node and destination itself, it can be encrypted by the applications using Tor – for example, the user's web browser could use HTTPS instead of HTTP.

At no single point in the circuit are the source IP address, destination IP address and contents of the message all known to an eavesdropper simultaneously. To reduce the chance that users can be de-anonymized, Tor attempts to avoid picking nodes that share the same operator when creating circuits. This makes it difficult, but perhaps not impossible, for the identity of a particular user to be discovered. For example, an attacker who can observe a user's traffic as it both enters and leaves the Tor network can carry out a traffic confirmation attack, in which they correlate characteristics such as the timing or volume of the user's traffic, to link the user to the destination server.

Unsurprisingly, the anonymity provided by Tor makes it an attractive tool for fraudsters. For example, a phisher who has tricked users into handing over their online banking credentials might use Tor to log in to the bank's website with the compromised credentials. The bank's log files will show the IP address of the Tor exit node, rather than the phisher's own IP address, making it more difficult for the bank and law enforcement agencies to trace the fraud back to the phisher.

The report from FinCEN examined 6,048 suspicious activity reports (SARs) filed by banks and other financial companies between 2001 and 2014. Of those, 975 involved Tor, totalling $24 million of "likely fraudulent activity". The report goes on to state that "in the majority of the SAR filings, the underlying suspicious activity – most frequently account takeovers – might have been prevented if the filing institution had been aware that their network was being accessed via Tor IP addresses." Even if blocking Tor does not deter phishers from committing fraud entirely, it may cause them to switch to using services that are easier for the authorities to trace, such as open proxy servers or anonymous VPN services.

According to FinCEN's report, banks were only aware that Tor was involved in 3% of cases. Netcraft has visited the websites of the ten financial companies most targeted by phishing in the last six months, using a variety of Tor exit nodes located around the world, to check if any of the companies block Tor.

Position Company Blocks Tor traffic
1 PayPal No, but Tor users must solve a CAPTCHA
3 AXA Banque No
4 SFR No
5 Wells Fargo No
6 Bank of America No
7 Chase No, but Tor users must use two-factor authentication
8 Lloyds Bank No
9 Banco do Brasil No
10 Cielo No

As shown in the table above, none of the login pages we visited blocked Tor traffic outright. For example, the following screenshot shows the appearance of PayPal's login page fetched from a variety of Tor exit nodes:

Screenshots of PayPal's login page fetched from several Tor exit nodes located across the world.

However, some of the websites we tested do treat Tor users differently during or after the login process – instead of blocking Tor users outright, they use Tor as an indicator for performing more stringent anti-fraud checks. (It is also possible that some companies perform additional checks that are not visible to end users.)

For example, Chase forces the use of two-factor authentication – by either email, text message or phone call – over Tor. PayPal requires Tor users to solve a CAPTCHA during the login process, which protects against automated attacks such as brute force login attempts, but would not prevent a phisher from manually logging into a victim's account. On the other hand, Lloyds Bank does not appear to visibly treat Tor users any differently to normal users.

A screenshot of the CAPTCHA that PayPal displays to users who attempt to log in over the Tor network.

The Tor Project considers services blanket blocking Tor traffic due to abusive and illegal behaviour by a proportion of its users to be a "threat to Tor's success". It advocates a range of other measures for sites to tackle abusive Tor traffic, including CAPTCHAs, two-factor authentication and establishing trust on a per-user rather than a per-IP basis. However, with the exception of two-factor authentication, most of these measures are targeted at abusive behaviour such as spam and are unlikely to prevent fraudsters from logging into compromised accounts.

Netcraft provides a wide range of countermeasures against phishing to many customers, including two of the world's top ten banks, as well as some smaller institutions at the sharp end of Internet crime – such as three of the largest Bitcoin exchanges and four Nigerian banks. For more information, please contact