The Anti-Phishing Working Group has released its latest Global Phishing Survey, written by myself and Rod Rasmussen. This report comprehensively examines a large data set of more than 250,000 confirmed phishing attacks detected in 2015 and 2016. By analyzing this cybercrime activity, we have learned more about what phishers have been doing, and how they have done it. Unfortunately, there’s more phishing than ever, and phishers are registering more domain names than ever.

Our major findings include:

1. In 2016, the number of phishing attacks, and the number of domain names used for phishing, reached an all-time high.

2. Malicious domain name registrations are also at an all-time high, indicating detection and mitigation problems at certain registrars and registries. Historically, most phishing has occurred on conpromised domains, where phishers broke into innocent registrants’ web hosting. But increasingly, phishers just go and register the domains they need. This major shift is concerning, and it means abuse detection and mitigation efforts are failing especially at certain registries and registrars.

3. Phishing in the new top-level domains (nTLDs) is rising and becoming more pervasive, but is not yet as pervasive as it is in the domain space as a whole. By the end of 2016, almost half of the nTLDs that were available for open registration had phishing in them. The nTLDs are also a place where phishers are purchasing domain names for themselves. Of the 6,549 domains used for phishing in the nTLDs in 2016, 86% (5,633) were registered maliciously.

4. New companies are constantly being targeted by phishers, while a few brands face an onslaught of thousands of attacks per year.

5. Contrary to conventional wisdom, phishers often wait up to three weeks before using domain names they have registered.

Full statistics and analysis of each of these topics—including breakdowns by TLDs and registrars—are included in the report.

In the meantime, phishers are employing another new trick that uses the domain name system. We call this “domain shadowing,” and is when a phisher manipulates an unsuspecting company’s DNS settings to insert multiple phishing sites onto the company’s servers. This often results in hundreds of new phishing sites at a time.

Our statistics under-count the total amount of phishing that occurred in the wild—more attacks were undetected by our sources, and more attacks were reported but not confirmed. The numbers are a baseline compiled through collection and counting methods that have remained consistent over the years.

Those who operate Internet resources have the responsibility to do so in a secure and wise manner. We hope this report is helpful and provides information that will make the Internet a better place.