In the field of computer security, phishing is the criminally fraudulent process of attempting to acquire sensitive information such as usernames, passwords and credit card details by masquerading as a trustworthy entity in an electronic communication. Communications purporting to be from popular social web sites, auction sites, online payment processors or IT administrators are commonly used to lure the unsuspecting public. Phishing is typically carried out by e-mail or instant messaging, and it often directs users to enter details at a fake website whose look and feel are almost identical to the legitimate one. Even when using server authentication, it may require tremendous skill to detect that the website is fake. Phishing is an example of social engineering techniques used to fool users, and exploits the poor usability of current web security technologies. Attempts to deal with the growing number of reported phishing incidents include legislation, user training, public awareness, and technical security measures.
A phishing technique was described in detail in 1987, and the first recorded use of the term "phishing" was made in 1996. The term is a variant of fishing, probably influenced by phreaking, and alludes to baits used to "catch" financial information and passwords.
A phishing technique was described in detail in 1987, in a paper and presentation delivered to the International HP Users Group, Interex. The first recorded mention of the term "phishing" is on the alt.online-service.America-online Usenet newsgroup on January 2, 1996, although the term may have appeared earlier in the print edition of the hacker magazine 2600.
Phishing on AOL was closely associated with the warez community that exchanged pirated software and the hacking scene that perpetrated credit card fraud and other online crimes. After AOL brought in measures in late 1995 to prevent using fake, algorithmically generated credit card numbers to open accounts, AOL crackers resorted to phishing for legitimate accounts and exploiting AOL.
A phisher might pose as an AOL staff member and send an instant message to a potential victim, asking him to reveal his password. In order to lure the victim into giving up sensitive information the message might include imperatives like "verify your account" or "confirm billing information". Once the victim had revealed the password, the attacker could access and use the victim's account for fraudulent purposes or spamming. Both phishing and warezing on AOL generally required custom-written programs, such as AOHell. Phishing became so prevalent on AOL that they added a line on all instant messages stating: "no one working at AOL will ask for your password or billing information", though even this didn't prevent some people from giving away their passwords and personal information if they read and believed the IM first. A user using both an AIM account and an AOL account from an ISP simultaneously could phish AOL members with relative impunity as internet AIM accounts could be used by non-AOL internet members and could not be actioned (ie- reported to AOL TOS department for disciplinary action.)
After 1997, AOL's policy enforcement with respect to phishing and warez became stricter and forced pirated software off AOL servers. AOL simultaneously developed a system to promptly deactivate accounts involved in phishing, often before the victims could respond. The shutting down of the warez scene on AOL caused most phishers to leave the service, and many phishers—often young teens—grew out of the habit.
The capture of AOL account information may have led phishers to misuse credit card information, and to the realization that attacks against online payment systems were feasible. The first known direct attempt against a payment system affected E-gold in June 2001, which was followed up by a "post-9/11 id check" shortly after the September 11 attacks on the World Trade Center. Both were viewed at the time as failures, but can now be seen as early experiments towards more fruitful attacks against mainstream banks. By 2004, phishing was recognized as a fully industrialized part of the economy of crime: specializations emerged on a global scale that provided components for cash, which were assembled into finished attacks.
Phishers are targeting the customers of banks and online payment services. E-mails, supposedly from the Internal Revenue Service, have been used to glean sensitive data from U.S. taxpayers. While the first such examples were sent indiscriminately in the expectation that some would be received by customers of a given bank or service, recent research has shown that phishers may in principle be able to determine which banks potential victims use, and target bogus e-mails accordingly. Targeted versions of phishing have been termed spear phishing. Several recent phishing attacks have been directed specifically at senior executives and other high profile targets within businesses, and the term whaling has been coined for these kinds of attacks.
Social networking sites are now a prime target of phishing, since the personal details in such sites can be used in identity theft; in late 2006 a computer worm took over pages on MySpace and altered links to direct surfers to websites designed to steal login details. Experiments show a success rate of over 70% for phishing attacks on social networks.
The RapidShare file sharing site has been targeted by phishing to obtain a premium account, which removes speed caps on downloads, auto-removal of uploads, waits on downloads, and cooldown times between downloads.
Attackers who broke into TD Ameritrade's database (containing all 6.3 million customers' social security numbers, account numbers and email addresses as well as their names, addresses, dates of birth, phone numbers and trading activity) also wanted the account usernames and passwords, so they launched a follow-up spear phishing attack.
Some people are being victimized by a Facebook Scam, the link being hosted by T35 Web Hosting and people are losing their accounts.
There are anti-phishing websites which publish exact messages that have been recently circulating the internet, such as FraudWatch International and Millersmiles. Such sites often provide specific details about the particular messages.
Most methods of phishing use some form of technical deception designed to make a link in an e-mail (and the spoofed website it leads to) appear to belong to the spoofed organization. Misspelled URLs or the use of subdomains are common tricks used by phishers. In the following example URL, http://www.yourbank.example.com/, it appears as though the URL will take you to the example section of the yourbank website; actually this URL points to the "yourbank" (i.e. phishing) section of the example website. Another common trick is to make the anchor text for a link appear to be valid, when the link actually goes to the phishers' site. The following example link, http://en.wikipedia.org/wiki/Genuine, appears to take you to an article entitled "Genuine"; clicking on it will in fact take you to the article entitled "Deception". In the lower left hand corner of most browsers you can preview and verify where the link is going to take you.
An old method of spoofing used links containing the '@' symbol, originally intended as a way to include a username and password (contrary to the standard). For example, the link http://firstname.lastname@example.org/ might deceive a casual observer into believing that it will open a page on www.google.com, whereas it actually directs the browser to a page on members.tripod.com, using a username of www.google.com: the page opens normally, regardless of the username supplied. Such URLs were disabled in Internet Explorer, while Mozilla Firefox and Opera present a warning message and give the option of continuing to the site or cancelling.
A further problem with URLs has been found in the handling of Internationalized domain names (IDN) in web browsers, that might allow visually identical web addresses to lead to different, possibly malicious, websites. Despite the publicity surrounding the flaw, known as IDN spoofing or homograph attack, phishers have taken advantage of a similar risk, using open URL redirectors on the websites of trusted organizations to disguise malicious URLs with a trusted domain. Even digital certificates do not solve this problem because it is quite possible for a phisher to purchase a valid certificate and subsequently change content to spoof a genuine website.
Phishers have used images instead of text to make it harder for anti-phishing filters to detect text commonly used in phishing e-mails.
An attacker can even use flaws in a trusted website's own scripts against the victim. These types of attacks (known as cross-site scripting) are particularly problematic, because they direct the user to sign in at their bank or service's own web page, where everything from the web address to the security certificates appears correct. In reality, the link to the website is crafted to carry out the attack, making it very difficult to spot without specialist knowledge. Just such a flaw was used in 2006 against PayPal.
A Universal Man-in-the-middle Phishing Kit, discovered by RSA Security, provides a simple-to-use interface that allows a phisher to convincingly reproduce websites and capture log-in details entered at the fake site.
To avoid anti-phishing techniques that scan websites for phishing-related text, phishers have begun to use Flash-based websites. These look much like the real website, but hide the text in a multimedia object.
Not all phishing attacks require a fake website. Messages that claimed to be from a bank told users to dial a phone number regarding problems with their bank accounts. Once the phone number (owned by the phisher, and provided by a Voice over IP service) was dialed, prompts told users to enter their account numbers and PIN. Vishing (voice phishing) sometimes uses fake caller-ID data to give the appearance that calls come from a trusted organization.
Another attack used successfully is to forward the client to a bank's legitimate website, then to place a popup window requesting credentials on top of the website in a way that it appears the bank is requesting this sensitive information.
The damage caused by phishing ranges from denial of access to e-mail to substantial financial loss. It is estimated that between May 2004 and May 2005, approximately 1.2 million computer users in the United States suffered losses caused by phishing, totaling approximately US$929 million. United States businesses lose an estimated US$2 billion per year as their clients become victims. In 2007 phishing attacks escalated. 3.6 million adults lost US $ 3.2 billion in the 12 months ending in August 2007. Microsoft claims these estimates are grossly exaggerated and puts the annual phishing loss in the US at $60 million. In the United Kingdom losses from web banking fraud—mostly from phishing—almost doubled to £23.2m in 2005, from £12.2m in 2004, while 1 in 20 computer users claimed to have lost out to phishing in 2005.
The stance adopted by the UK banking body APACS is that "customers must also take sensible precautions ... so that they are not vulnerable to the criminal." Similarly, when the first spate of phishing attacks hit the Irish Republic's banking sector in September 2006, the Bank of Ireland initially refused to cover losses suffered by its customers (and it still insists that its policy is not to do so), although losses to the tune of €11,300 were made good.
There are several different techniques to combat phishing, including legislation and technology created specifically to protect against phishing.
One strategy for combating phishing is to train people to recognize phishing attempts, and to deal with them. Education can be effective, especially where training provides direct feedback. One newer phishing tactic, which uses phishing e-mails targeted at a specific company, known as spear phishing, has been harnessed to train individuals at various locations, including United States Military Academy at West Point, NY. In a June 2004 experiment with spear phishing, 80% of 500 West Point cadets who were sent a fake e-mail were tricked into revealing personal information.
People can take steps to avoid phishing attempts by slightly modifying their browsing habits. When contacted about an account needing to be "verified" (or any other topic used by phishers), it is a sensible precaution to contact the company from which the e-mail apparently originates to check that the e-mail is legitimate. Alternatively, the address that the individual knows is the company's genuine website can be typed into the address bar of the browser, rather than trusting any hyperlinks in the suspected phishing message.
Nearly all legitimate e-mail messages from companies to their customers contain an item of information that is not readily available to phishers. Some companies, for example PayPal, always address their customers by their username in e-mails, so if an e-mail addresses the recipient in a generic fashion ("Dear PayPal customer") it is likely to be an attempt at phishing. E-mails from banks and credit card companies often include partial account numbers. However, recent research has shown that the public do not typically distinguish between the first few digits and the last few digits of an account number—a significant problem since the first few digits are often the same for all clients of a financial institution. People can be trained to have their suspicion aroused if the message does not contain any specific personal information. Phishing attempts in early 2006, however, used personalized information, which makes it unsafe to assume that the presence of personal information alone guarantees that a message is legitimate. Furthermore, another recent study concluded in part that the presence of personal information does not significantly affect the success rate of phishing attacks, which suggests that most people do not pay attention to such details.
The Anti-Phishing Working Group, an industry and law enforcement association, has suggested that conventional phishing techniques could become obsolete in the future as people are increasingly aware of the social engineering techniques used by phishers. They predict that pharming and other uses of malware will become more common tools for stealing information.
Everyone can help educate the public by encouraging safe practices, and by avoiding dangerous ones. Unfortunately, even well-known players are known to incite users to hazardous behaviour, e.g. by requesting their users to reveal their passwords for third party services, such as email.
Anti-phishing measures have been implemented as features embedded in browsers, as extensions or toolbars for browsers, and as part of website login procedures. The following are some of the main approaches to the problem.
Most websites targeted for phishing are secure websites, meaning that SSL with strong cryptography is used for server authentication, where the website's URL is used as identifier. In theory it should be possible for the SSL authentication to be used to confirm the site to the user, and this was SSL v2's design requirement and the meta of secure browsing. But in practice, this is easy to trick.
The superficial flaw is that the browser's security user interface (UI) is insufficient to deal with today's strong threats. There are three parts to secure authentication using TLS and certificates: indicating that the connection is in authenticated mode, indicating which site the user is connected to, and indicating which authority says it is this site. All three are necessary for authentication, and need to be confirmed by/to the user.
Secure Connection. The standard display for secure browsing from the mid-1990s to mid-2000s was the padlock. In 2005, Mozilla fielded a yellow URL bar 2005 as a better indication of the secure connection. This innovation was later reversed due to the EV certificates, which replaced certain certificates providing a high level of organization identity verification with a green display, and other certificates with an extended blue favicon box to the left of the URL bar (in addition to the switch from "http" to "https" in the url itself).
Which Site. The user is expected to confirm that the domain name in the browser's URL bar was in fact where they intended to go. URLs can be too complex to be easily parsed. Users often do not know or recognise the URL of the legitimate sites they intend to connect to, so that the authentication becomes meaningless. A condition for meaningful server authentication is to have a server identifier that is meaningful to the user; many ecommerce sites will change the domain names within their overall set of websites, adding to the opportunity for confusion. Simply displaying the domain name for the visited website as some anti-phishing toolbars do is not sufficient.
Some newer browsers display the entire URL in grey, with just the domain name itself in black, as a means of assisting users in identifying fraudulent URLs.
An alternate approach is the petname extension for Firefox which lets users type in their own labels for websites, so they can later recognize when they have returned to the site. If the site is not recognised, then the software may either warn the user or block the site outright. This represents user-centric identity management of server identities. Some suggest that a graphical image selected by the user is better than a petname.
With the advent of EV certificates, browsers now typically display the organisation's name in green, which is much more visible and is hopefully more consistent with the user's expectations. Unfortunately, browser vendors have chosen to limit this prominent display only to EV certificates, leaving the user to fend for himself with all other certificates.
Who is the Authority. The browser needs to state who the authority is that makes the claim of who the user is connected to. At the simplest level, no authority is stated, and therefore the browser is the authority, as far as the user is concerned. The browser vendors take on this responsibility by controlling a root list of acceptable CAs. This is the current standard practice.
The problem with this is that not all certification authorities (CAs) employ equally good nor applicable checking, regardless of attempts by browser vendors to control the quality. Nor do all CAs subscribe to the same model and concept that certificates are only about authenticating ecommerce organisations. Certificate Manufacturing is the name given to low-value certificates that are delivered on a credit card and an email confirmation; both of these are easily perverted by fraudsters. Hence, a high-value site may be easily spoofed by a valid certificate provided by another CA. This could be because the CA is in another part of the world, and is unfamiliar with high-value ecommerce sites, or it could be that no care is taken at all. As the CA is only charged with protecting its own customers, and not the customers of other CAs, this flaw is inherent in the model.
The solution to this is that the browser should show, and the user should be familiar with, the name of the authority. This presents the CA as a brand, and allows the user to learn the handful of CAs that she is likely to come into contact within her country and her sector. The use of brand is also critical to providing the CA with an incentive to improve their checking, as the user will learn the brand and demand good checking for high-value sites.
This solution was first put into practice in early IE7 versions, when displaying EV certificates. In that display, the issuing CA is displayed. This was an isolated case, however. There is resistance to CAs being branded on the chrome, resulting in a fallback to the simplest level above: the browser is the user's authority.
Experiments to improve the security UI have resulted in benefits, but have also exposed fundamental flaws in the security model. The underlying causes for the failure of the SSL authentication to be employed properly in secure browsing are many and intertwined.
Security before threat. Because secure browsing was put into place before any threat was evident, the security display lost out in the "real estate wars" of the early browsers. The original design of Netscape's browser included a prominent display of the name of the site and the CA's name, but these were dropped in the first release. Users are now highly experienced in not checking security information at all.
Click-thru syndrome. However, warnings to poorly configured sites continued, and were not down-graded. If a certificate had an error in it (mismatched domain name, expiry), then the browser would commonly launch a popup to warn the user. As the reason was generally misconfiguration, the users learned to bypass the warnings, and now, users are accustomed to treat all warnings with the same disdain, resulting in Click-thru syndrome. For example, Firefox 3 has a 4-click process for adding an exception, but it has been shown to be ignored by an experienced user in a real case of MITM. Even today, as the vast majority of warnings will be for misconfigurations not real MITMs, it is hard to see how click-thru syndrome will ever be avoided.
Lack of interest. Another underlying factor is the lack of support for virtual hosting. The specific causes are a lack of support for Server Name Indication in TLS webservers, and the expense and inconvenience of acquiring certificates. The result is that the use of authentication is too rare to be anything but a special case. This has caused a general lack of knowledge and resources in authentication within TLS, which in turn has meant that the attempts by browser vendors to upgrade their security UIs have been slow and lacklustre.
Lateral communications. The security model for secure browser includes many participants: user, browser vendor, developers, CA, auditor, webserver vendor, ecommerce site, regulators (e.g., FDIC), and security standards committees. There is a lack of communication between different groups that are committed to the security model. E.g., although the understanding of authentication is strong at the protocol level of the IETF committees, this message does not reach the UI groups. Webserver vendors do not prioritise the Server Name Indication (TLS/SNI) fix, not seeing it as a security fix but instead a new feature. In practice, all participants look to the others as the source of the failures leading to phishing, hence the local fixes are not prioritised.
Matters improved slightly with the CAB Forum, as that group includes browser vendors, auditors and CAs. But the group did not start out in an open fashion, and the result suffered from commercial interests of the first players, as well as a lack of parity between the participants. Even today, CAB forum is not open, and does not include representation from small CAs, end-users, ecommerce owners, etc.
Standards gridlock. Vendors commit to standards, which results in an outsourcing effect when it comes to security. Although there have been many and good experiments in improving the security UI, these have not been adopted because they are not standard, or clash with the standards. Threat models can re-invent themselves in around a month; Security standards take around 10 years to adjust.
Venerable CA model. Control mechanisms employed by the browser vendors over the CAs have not been substantially updated; the threat model has. The control and quality process over CAs is insufficiently tuned to the protection of users and the addressing of actual and current threats. Audit processes are in great need of updating. The recent EV Guidelines documented the current model in greater detail, and established a good benchmark, but did not push for any substantial changes to be made.
Another popular approach to fighting phishing is to maintain a list of known phishing sites and to check websites against the list. Microsoft's IE7 browser, Mozilla Firefox 2.0, Safari 3.2, and Opera all contain this type of anti-phishing measure. Firefox 2 used Google anti-phishing software. Opera 9.1 uses live blacklists from PhishTank and GeoTrust, as well as live whitelists from GeoTrust. Some implementations of this approach send the visited URLs to a central service to be checked, which has raised concerns about privacy. According to a report by Mozilla in late 2006, Firefox 2 was found to be more effective than Internet Explorer 7 at detecting fraudulent sites in a study by an independent software testing company.
An approach introduced in mid-2006 involves switching to a special DNS service that filters out known phishing domains: this will work with any browser, and is similar in principle to using a hosts file to block web adverts.
To mitigate the problem of phishing sites impersonating a victim site by embedding its images (such as logos), several site owners have altered the images to send a message to the visitor that a site may be fraudulent. The image may be moved to a new filename and the original permanently replaced, or a server can detect that the image was not requested as part of normal browsing, and instead send a warning image.
The Bank of America's website is one of several that ask users to select a personal image, and display this user-selected image with any forms that request a password. Users of the bank's online services are instructed to enter a password only when they see the image they selected. However, a recent study suggests few users refrain from entering their password when images are absent. In addition, this feature (like other forms of two-factor authentication) is susceptible to other attacks, such as those suffered by Scandinavian bank Nordea in late 2005, and Citibank in 2006.
A similar system, in which an automatically-generated “Identity Cue” consisting of a colored word within a colored box is displayed to each website user, is in use at other financial institutions.
Security skins are a related technique that involves overlaying a user-selected image onto the login form as a visual cue that the form is legitimate. Unlike the website-based image schemes, however, the image itself is shared only between the user and the browser, and not between the user and the website. The scheme also relies on a mutual authentication protocol, which makes it less vulnerable to attacks that affect user-only authentication schemes.
Specialized spam filters can reduce the number of phishing e-mails that reach their addressees' inboxes. These approaches rely on machine learning and natural language processing approaches to classify phishing e-mails.
Several companies offer banks and other organizations likely to suffer from phishing scams round-the-clock services to monitor, analyze and assist in shutting down phishing websites. Individuals can contribute by reporting phishing to both volunteer and industry groups, such as PhishTank. Individuals can also contribute by reporting phone phishing attempts to Phone Phishing, Federal Trade Commission.
These protocols enable weak passwords to provide strong security even with a man in the middle. The password is never sent to the server, nor is any information that would enable a hostile server to perform a brute force attack. They thus effectively make phishing impossible, without any need of PKI certificates, user warnings or other mechanisms.
However, as of Jan 2010 they are not implemented in any widely used browsers, although this may change. If they were widely implemented and used they could damage the substantial anti-phishing industry.
On January 26, 2004, the U.S. Federal Trade Commission filed the first lawsuit against a suspected phisher. The defendant, a Californian teenager, allegedly created a webpage designed to look like the America Online website, and used it to steal credit card information. Other countries have followed this lead by tracing and arresting phishers. A phishing kingpin, Valdir Paulo de Almeida, was arrested in Brazil for leading one of the largest phishing crime rings, which in two years stole between US$18 million and US$37 million. UK authorities jailed two men in June 2005 for their role in a phishing scam, in a case connected to the U.S. Secret Service Operation Firewall, which targeted notorious "carder" websites. In 2006 eight people were arrested by Japanese police on suspicion of phishing fraud by creating bogus Yahoo Japan Web sites, netting themselves 100 million yen ($870,000 USD). The arrests continued in 2006 with the FBI Operation Cardkeeper detaining a gang of sixteen in the U.S. and Europe.
In the United States, Senator Patrick Leahy introduced the Anti-Phishing Act of 2005 in Congress on March 1, 2005. This bill, if it had been enacted into law, would have subjected criminals who created fake web sites and sent bogus e-mails in order to defraud consumers to fines of up to $250,000 and prison terms of up to five years. The UK strengthened its legal arsenal against phishing with the Fraud Act 2006, which introduces a general offence of fraud that can carry up to a ten year prison sentence, and prohibits the development or possession of phishing kits with intent to commit fraud.
Companies have also joined the effort to crack down on phishing. On March 31, 2005, Microsoft filed 117 federal lawsuits in the U.S. District Court for the Western District of Washington. The lawsuits accuse "John Doe" defendants of obtaining passwords and confidential information. March 2005 also saw a partnership between Microsoft and the Australian government teaching law enforcement officials how to combat various cyber crimes, including phishing. Microsoft announced a planned further 100 lawsuits outside the U.S. in March 2006, followed by the commencement, as of November 2006, of 129 lawsuits mixing criminal and civil actions. AOL reinforced its efforts against phishing in early 2006 with three lawsuits seeking a total of $18 million USD under the 2005 amendments to the Virginia Computer Crimes Act, and Earthlink has joined in by helping to identify six men subsequently charged with phishing fraud in Connecticut.
In January 2007, Jeffrey Brett Goodin of California became the first defendant convicted by a jury under the provisions of the CAN-SPAM Act of 2003. He was found guilty of sending thousands of e-mails to America Online users, while posing as AOL's billing department, which prompted customers to submit personal and credit card information. Facing a possible 101 years in prison for the CAN-SPAM violation and ten other counts including wire fraud, the unauthorized use of credit cards, and the misuse of AOL's trademark, he was sentenced to serve 70 months. Goodin had been in custody since failing to appear for an earlier court hearing and began serving his prison term immediately.
[[File:|right|thumbnail|300px|An example of a phishing email, disguised as an official email from a (fictional) bank. The sender is attempting to trick the recipient into revealing secure information by "confirming" it at the phisher's website.]] Phishing is a way that criminals get sensitive information (like usernames or passwords). It is a method of social engineering. Very often, phishing is done by electronic mail. This mail appears to come from a bank or other service provider. It usually says that because of some change in the system, the users need to re-enter their usernames/passwords to confirm them. The emails usually have a link to a page that looks almost like that of the real bank.
Phishing grew by 42% in January of 2005: 8,829 E-mails with phishing were sent in December of 2004 but that grew to 12,845 by end of January 2005.
The number of web pages with phishing grew from 1,740 to 2,560 in the same period.