January 2010 Report #27 The data in this report is aggregated from a combination of sources including Symantec’s Phish Report Network (PRN), strategic partners, customers and security solutions. This report discusses the metrics and trends observed in phishing activity during the month of December 2009. Highlighted in the January 2010 report: Symantec observed a 4 percent decrease from the previous month in all phishing at- tacks 21 percent of phishing URLs were generated using phishing toolkits; a decrease of 19 percent from the previous month A 26 percent decrease from the previous month was observed in non-English phishing sites More than 118 Web hosting services were used, which accounted for 11 percent of all phishing attacks; an increase of 2 percent in total Web host URLs when compared to the previous month Phishing Tactic Distribution: Phishing sites were categorized based upon the domains they leveraged. In December, the total volume of phishing decreased by 4 percent from the previous month. Toolkit based phishing attacks faced a decrease of 19 percent, which most likely led to the decrease in the overall volume of phishing. However, phishing attacks based on other methods including typosquatting, IP domains and use of webhosting sites have slightly increased from the previous month. David Cowings Mathew Maniyara Sagar Desai Executive Editor Editor PR Contact Security Response Security Response Sagar_desai@symantec.com Phishing site attack methods and target sectors Phishing sites in December 2009 were categorized to understand the attack methods and determine the sectors and brands impacted by the attacks. The following categories were analyzed: Sectors Number of brands Phishing toolkits Fraud URLs with IP addresses Phish sites that use IP address domains – categorized by hosted cities Use of Web hosting sites Geo-locations of phishing sites Non-English phishing sites Top-Level domains of phishing sites Country of brand Non-English phishing sites Sectors: Phishing target sectors are seen in the graphic below. Number of Brands: Symantec observed that 79 percent of all attacks were from unique phishing websites, which included more than 218 targeted brands. In December 2009, the unique phishing activity increased by 2 percent over the previous month. The proportion of unique phishing URLs increased from 75 percent (in November 2009) to 79 percent (in December 2009). The higher proportion of unique phishing activity was a result of a fall in the number of toolkit phishing websites in the month. Weekly Behavior of Phishing Toolkit Activity Automated Phishing Toolkits: Symantec observed that 21 percent of rest of the month. The toolkit attacks on the phishing URLs were generated using phishing financial sector decreased considerably in De- toolkits. The number of toolkit attacks cember 2009. The information services sector decreased considerably by 19 percent. A had the majority of toolkit attacks during the spike was observed in the first week of the month. The attacks on the financial sector month in toolkit phishing activity. However, spanned several brands whereas the attacks the volume of phishing attacks from these in information services sector were primarily toolkits was considerably low throughout the targeted at a single brand. Phishing Attacks Using IP Address Domains Phishers today use IP addresses as part of the hostname instead of a domain name. This is a tactic employed to hide the actual fake domain name that otherwise can easily be noticed. Also, many banks use IP addresses in their website URLs. A total of 1135 phishing sites were hosted in 61 countries. This amounted to an increase of approximately 8 percent of IP attacks in comparison to the previous month. The United States continued to be the top ranked country hosting phishing sites. The Greater China region con- tinued to be at the second spot comprising of 7 percent of IP attacks. South Korea accounted for approximately 7 percent of IP attacks in the month. The top cities hosting phish sites were Seoul, Fort Lauderdale and Clarksville. Debuting in December was Clarksville at the third position. Brooklyn, Atlanta and Montreal were common to the list of top cities in the previous month as well. Phishing Exploits of Free Web Hosting Services For phishers, using free web hosting services has been the easiest form of phishing in terms of cost and technical skills required to develop fake sites. A total of 118 different web hosting services However, this form of attack is not as widely served as the home for 2,150 phishing sites in used as it frequently requires manual efforts the month of December. Symantec observed to prepare the phishing Web page, unlike the a 2 percent increase in the number of free automated kit generated websites. Many free web hosting services utilized for developing web hosts have also improved their preventa- phishing sites. More than 84 brands were tive and corrective anti-phishing measures attacked using this method in the reporting significantly decreasing the lifespan of phish- period. ing sites on their systems. Global Distribution of Phishing Sites Phishing sites were analyzed based upon the geo-location of their web hosts as well as the number of unique URL’s (referred as “lures” in the report) utilized to lure victims to the phish- 1. Geo-Location of Phishing Lures Leading this area are the USA (37 percent), in the previous month to (37 percent) in the South Korea (5 percent) and Canada (5 per- current month. The number of active lures cent). The proportion of lures from the USA from Canada exceeded that from Germany has decreased considerably from (43 percent) and has taken over the third position. 2. Geo-Location of Phishing Web Hosts The top countries are the USA (43 percent), previous month. In December, the distribu- Germany (4 percent) and South Korea (3 per- tion of web hosts was evenly distributed for cent). Symantec observed that the phishing all other locations. hosts originated from more countries than the Geo-Location of Phishing Web Hosts Non-English Phishing Trends Phishing attacks in Italian, French and Portu- guese languages were the highest in Non- English language attacks. In December 2009, the number of attacks in Portuguese in- creased further, exceeding Chinese and reach- ing the third position. The increase in phishing on a Brazilian social networking website was the reason behind the increase of attacks in Portuguese language. Phishing websites in Italian and French remained higher in financial brands. The attacks in Chinese language pre- vailed in the e-commerce sector. Top-Level Domains of Phishing Sites Phishing URLs were categorized based on the Top-Level Domains (TLD). TLDs are the last part of an Internet domain name; i.e., the letters that follow the final dot of any domain name. E.g., in the domain name www.example.com, the Top-Level Domain is .com (or COM, as domain names are not case-sensitive). Country Code Top-Level Domains (ccTLD) are used by a country or a territory. They are two letters long, for example .us is for the United States. Generic Top- Level Domains (gTLD) are used by a particular type of organization (.com for a commercial organization). It is three or more letters long. Most gTLDs are available for use worldwide, but for historical reasons .mil (military) and .gov (government) are restricted to use by the respec- tive U.S. authorities. Comparisons of Top-Level Domains of Phishing Sites Overall TLDs The most used TLDs in phishing sites in the month of December were, .com, .net and .org com- prising of (54 percent), (7 percent) and (4 percent) respectively. The Top-Level Domains in phishing were then further categorized: 1. Generic Top-Level Domains (gTLDs) The generic TLDs .com, .net and .co were the most utilized with (74 percent), (10 percent) and (6 percent) of the total phish attacks respectively. 2. Country Code Top-Level Domains (ccTLDs) The Russian, Chinese and United Kingdom ccTLDs were evaluated to be the highest in phishing attacks with (11 percent), (8 percent) and (7 percent) respectively. Country of Targeted Brands The brands that phishing sites spoofed were categorized based on the country in which the brand’s parent company is based. The top countries of brands attacked in December were the USA, UK and Italy. There were 32 countries whose brands were attacked. As seen in the previous months, the trend of the sec- tors targeted is similar throughout the countries of brand origin except for those belonging to Germany and China. There was a combination of banking, e-commerce and information ser- vices sectors in German brands. In China, the e-commerce sector remains a primary target. UAE made its debut in the top countries of brands attacked. A higher number of phishing attacks on a bank based in UAE led to the country making its debut in the list. Another contributing factor was the disappearance of France and Israel from the list. There were fewer phishing attacks on brands based in these two countries as compared to the previous month. Glossary of Terms Phishing Toolkits: Phishing toolkits are auto- Top-Level Domain (TLD): Sometimes referred mated toolkits that facilitate the creation of to as a Top-Level Domain Name (TLDN): It is phishing Websites. They allow individuals to the last part of an Internet domain name; that create and carry out phishing attacks even is, the letters that follow the final dot of any without any technical knowledge. domain name. For example, in the domain name www.example.com, the Top-Level Do- Unique Phishing Web site: The phishing Web main is com (or COM, as domain names are sites that have a unique Web page are classi- not case-sensitive). fied as “Unique Phishing Websites”. URLs from phishing toolkits that randomize their Country Code Top-Level Domains (ccTLD): URL string are observed to point to the same Used by a country or a dependent territory. It Web page and do not contain a unique Web is two letters long, for example .us for the page in each URL. Unique Phishing Web sites United States. are the ones where each attack is categorized on distinct Web pages. Generic Top-Level Domains (gTLD): Used by a particular class of organizations (for example, Web-Hosting: Type of Internet hosting ser- .com for commercial organizations). It is vice which allows individuals and organiza- three or more letters long. Most gTLDs are tions to put up their own websites. These available for use worldwide, but for historical websites run on the space of Web host com- reasons .mil (military) and .gov pany servers accessible via the World Wide (governmental) are restricted to use by the Web. There are different types of Web host- respective U.S. Authorities. gTLDs are sub ing services namely, free Web hosting, shared classified into sponsored Top-Level Domains Web hosting, dedicated Web hosting, man- (sTLD), e.g. .aero, .coop and .museum, and un- aged Web hosting, etc. of which the free Web sponsored Top-Level Domains (uTLD), e.g. hosting service is commonly used to create .biz, .info, .name and .pro. phishing websites. Typo-Squatting: Typo-squatting refers to the practice of registering domain names that are typo variations of financial institution web- sites or other popular websites. Phishing Lure: Phishing lures are URLs distrib- uted in spam/phishing email utilized to lure victims to fraudulent phishing websites.