Research finds homograph attacks on big brands rife

Apparent domain name homograph attacks against major brands are a “significant” problem, according to research from Farsight Security.
The company said last week that it scanned for such attacks against 125 well-known brands over the three months to January 10 and found 116,113 domains — almost 1,000 per brand.
Homographs are domains that look like other domains, often indistinguishable from the original. They’re usually used to phish for passwords to bank accounts, retailers, cryptocurrency exchanges, and so on.
They most often use internationalized domain names, mixing together ASCII and non-ASCII characters when displayed in browsers.
To the naked eye, they can look very similar to the original ASCII-only domains, but under the hood they’re actually encoded with Punycode with the xn-- prefix.
Examples highlighted by Farsight include baŋkofamerica.com, amazoṇ.com and fàcebook.com
Displayed as ASCII, those domains are actually xn--bakofamerica-qfc.com, xn--amazo-7l1b.com and xn--fcebook-8va.com.
Farsight gave examples including and excluding the www. subdomain in a blog post last week, but I’m not sure if it double-counted to get to its 116,113-domain total.
As you might imagine, almost all of this abuse is concentrated in .com and other TLDs that were around before 2012, judging by Farsight’s examples. That’s because the big brands are not using new gTLDs for their primary sites yet.
Farsight gave a caveat that it had not generally investigated the ownership of the homograph domains it found. It’s possible some of them are defensive registrations by brands that are already fully aware of the security risk they could present.

Comment Tagged: farsight, homograph, idn, phishing, security, trademarks

DomainTools scraps apps and APIs in war on spam

DomainTools is to scrap at least five of its services as it tries to crack down spam.
It’s getting rids of its mobile apps, its APIs, and is to stop showing registrants’ personal information to unauthenticated users.
CEO Tim Chen told us in an email at the weekend:

The Android app is no longer supported.
The iOS app will no longer be supported after February 20th.
The Developer API is no longer supported.
On February 20th, the Bulk Parsed Whois tool available to Personal Members will no longer be supported.
On February 20th, our production Whois API will no longer be available to individual membership levels, an Enterprise relationships will be required.

It’s all part of an effort to make sure DomainTools services are not being abused by spammers, which has lead to a dispute with GoDaddy over bulk access to its registrants’ Whois data.
The longstanding problem of new registrants getting spammed with calls and emails offering web hosting and such has escalated over the last few years. Domain Name Wire detailed the scale of the abuse registrants can experience in a post last week.
While to my knowledge nobody has directly accused DomainTools of facilitating such abuse, the scrapped services are the ones that would be most useful to these spammers.
The company is also going to scale back what guest users can see when they do a Whois lookup, and is to make automated scraping of Whois records more difficult for paying members.
In a blog post, Chen wrote last week:

As of today, unauthenticated users of the DomainTools Whois Lookup tool will not see personally identifiable information for the registrant parsed out in the results, and will be required to submit a CAPTCHA to see the full raw domain name Whois record. Phone numbers in the parsed results have been replaced with image files, much the same way emails have always been rendered

As well as hoping to ease relations with GoDaddy — the source of a very heavy chunk of DomainTools’ data — the moves are also part of the company’s strategy for dealing with the incoming General Data Protection Regulation.
This is the EU law that gives registrants more control over the privacy of their personal data.
Chen told us earlier this month that DomainTools is keen to ensure its enterprise-level suite of security products, which he said are vital for security and intellectual property investigations, continue to operatie under the new regime.
About 80% of DomainTools’ revenue comes from its enterprise-level customers, over 500 companies.

2 Comments Tagged: .app, api, DomainTools, gdpr, godaddy, privacy, whois

A new gTLD kills itself off for the second time

British pharmacy chain Boots has applied to ICANN to terminate its dot-brand contract for the second time.
The company asked for its .boots Registry Agreement, signed in 2015, to be ended in December and ICANN opened the request for public comment this week.
What’s weird about the request is that Boots had already asked for self-termination last April, but that request was subsequently withdrawn by the company.
Boots seems to have changed its mind, twice, in a year.
As I noted first time around, .boots was the first example of a dot-brand that also matches a generic class of goods to chose the easy way out.
It’s quite likely the two-year freeze on re-applying for the string, should anyone want to, will be over by the time the next new gTLD application window opens.
.boots only had the contractually mandated placeholder domain nic.boots live.

3 Comments Tagged: boots, dot-brands, ICANN, new gTLDs, termination

ICANN blocks 1.5 million domains, including some three-letter names

A million and a half domain names, including many potential valuable three and four-letter strings, have been been given special protection across all gTLDs under a new ICANN policy.
The long-discussed, highly controversial reservation of the names and acronyms of various intergovernmental and non-governmental organizations has become official ICANN Consensus Policy and will be binding on all gTLD registries and registrars from August this year.
The policy gives special protection to (by my count) 1,282 strings in each of the (again, by my count) 1,243 existing gTLDs, as well as future gTLDs. That comes to over 1.5 million domains.
The strings match the names, and sometimes the acronyms and abbreviations, of recognized Intergovernmental Organizations (IGOs) and International Non-Governmental Organizations (INGOs) as well as the International Olympic Committee, Red Cross, Red Crescent and related movements.
These are all organizations whose names are protected by international law but not necessarily by trademarks.
Protected strings run from obscurities such as “europeanbankforreconstructionanddevelopment” and “internationalunionfortheprotectionofnewvarietiesofplants” to “can”, “eco” and “fao”.
All gTLDs, including legacy TLDs such as .com, are affected by the policy.
The full list of protected strings can be found here.
Any of the Red Cross, IOC and IGO strings already registered will remain registered, and registries are obliged to honor renewal and transfer requests. Nobody’s losing their domains, in other words. But if any are deleted, they must be clawed back and reserved by the registry.
The protected organizations must be given the ability to register their reserved matching names should they wish to, the policy states.
Registries will be able to sell the acronyms of protected INGOs, but will have to offer an “INGO Claims Service”, which mirrors the existing Trademark Claims service, in gTLDs that go live in future.
The policy was developed by ICANN’s Generic Names Supporting Organization and approved by the ICANN board of directors all the way back in April 2014 and has been in implementation talks ever since.
It’s the 14th Consensus Policy to be added to ICANN’s statute book since the organization was formed 20 year ago.
Registries and registrars have until August 1 to make sure they’re compliant. Consensus Policies are basically incorporated into their contracts by reference.
Work on IGO/INGO protections is actually still ongoing. There’s a GNSO Policy Development Process on “curative” rights for IGOs and INGOs (think: UDRP) that is fairly close to finishing its work but is currently mired in a mind-numbing process debate.
UPDATE: This post was updated January 17, 2018 to correct the number of reserved strings and to clarify how INGO names are treated by the policy.

5 Comments Tagged: consensus policy, ICANN, igo, ingo, ioc, olympic, red cross, reserved

Three ways ICANN could gut Whois

ICANN has published three possible models of how Whois could be altered beyond recognition after European privacy law kicks in this May.
Under each model, casual Whois users would no longer have access to the wealth of contact information they do under the current system.
There may also be a new certification program that would grant access to full Whois records to law enforcement, consumer protection agencies and intellectual property interests.
The three models are each intended to address the General Data Protection Regulation, EU law that could see companies fined millions if they fail to protect the personal data of European citizens.
While GDPR affects all data collection on private citizens, for the domain name industry it’s particularly relevant to Whois, where privacy has always been an afterthought.
The three ICANN models, which are now subject to a short public comment period, differ from each other in three key areas: who has their privacy protected, which fields appear in public Whois by default, and how third parties such as law enforcement access the full records.
Model 1 is the most similar to the current system, allowing for the publication of the most data.
Under this model the name and postal address of the registrant would continue to be displayed in the public Whois databases.
Their email address and phone number would be protected, but the email and phone of the administrative and technical contacts — often the same person as the registrant — would be published.
If the registrant were a legal entity, rather than a person, all data fields would continue to be displayed as normal.
The other two models call for more restricted, or at least different, public output.
Under Model 2, the email addresses of the administrative and technical contacts would be published, but all other contact information, including the name of the registrant, would be redacted.
Model 3 proposes a crazy-sounding system whereby everything would be published unless the registrar/registry decided, on a domain-by-domain basis, that the field contained personal information.
This would require manual vetting of each Whois record and is likely to gather no support from the industry.
The three models also differ in how third parties with legitimate interests would access full Whois records.
Model 1 proposes a system similar to how zone files are published via ICANN’s Centralized Zone Data Service.
Under this model, users would self-certify that they have a legit right to the data (if they’re a cop or an IP lawyer, for example) and it would be up to the registry or registrar to approve or decline their request.
Model 2 envisages a more structured, formal, centralized system of certification for Whois users, developed with the Governmental Advisory Committee and presumably administered by ICANN.
Model 3 would require Whois users to supply a subpoena or court order in order to access records, which is sure to make it unpopular among the IP lobby and governments.
Each of the three models also differs in terms of the circumstances under which privacy is provided.
The models range from protecting records only when the registrant, registry, registrar or any other entity involved in the data processing has a presence in the European Economic Area to protecting records of all registrants everywhere regardless of whether they’re a person or a company.
Each model has different data retention policies, ranging from six month to two years after a registration expires.
None of the three models screw with registrars’ ability to pass data to thick-Whois registries, nor to their data escrow providers.
ICANN said it’s created these models based on the legal analyses it commissioned from the Hamilton law firm, as well as submissions from community members.
One such submission, penned by the German trade associated Eco, has received broad industry support.
It would provide blanket protection to all registrants regardless of legal status or location, and would see all personally identifiable information stripped from public Whois output.
Upon carrying out a Whois query, users would see only information about the domain, not the registrant.
There would be an option to request more information, but this would be limited to an anonymized email address or web form for most users.
Special users, such as validated law enforcement or IP interests, would be able to access the full records via a new, centralized Trusted Data Clearinghouse, which ICANN would presumably be responsible for setting up.
It’s most similar to ICANN’s Model 2.
It has been signed off by registries and registrars together responsible for the majority of the internet’s domain registrations: Afilias, dotBERLIN, CentralNic, Donuts, Neustar, Nominet, Public Interest Registry (PIR), Verisign, 1&1, Arsys, Blacknight, GoDaddy, Strato/Cronon, Tucows and United Domains.
ICANN said in a blog post that its three models are now open for public comment until January 29.
If you have strong opinions on any of the proposals, it might be a good idea to get them in as soon as possible, because ICANN plans to identify one of the models as the basis for the official model within 48 hours of the comment period closing.

2 Comments Tagged: .eco, gdpr, hamilton, ICANN, privacy, whois

Ramchandani promoted to Radix CEO

New gTLD registry Radix has appointed long-time business head Sandeep Ramchandani as CEO.
He’s replacing Bhavin Turakhia, who is CEO of parent company Directi and executive chairman of Radix.
Ramchandani had a lot of autonomy as business head and VP of the company and, in my view, has been basically CEO in all but name for years. I’ve accidentally called him CEO in the pages of DI more than once.
In a press release, he said: “Just as the first few years of Radix were about demonstrating proof of concept, the next few will be about growing awareness and delivering accelerated growth. We are also actively looking to acquire more TLD assets to reach newer segments of the market while leveraging economies of scale.”
The company has a portfolio of nine gTLDs, including .website, .store and .online, and recently announced that its 2017 revenue topped $12 million.

Comment Tagged: directi, radix

GoDaddy and DomainTools scrap over Whois access

GoDaddy has seriously limited DomainTools’ access to its customers’ Whois records, pissing off DomainTools.
DomainTools CEO Tim Chen this week complained to DI that its access to Whois has been throttled back significantly in recent months, making it very difficult to keep its massive database of domain information up to date.
Chen said that DomainTools is currently only able to access GoDaddy’s Whois over port 43 at about 2% of the rate it had previously.
He said that this has been going on for about six months and that the market-leading registrar has been unresponsive to its requests to have previous levels restored.
“By throttling access to the data by 98% they’re defeating the ability of security practitioners to get data on GoDaddy domains,” Chen said. “It’s particularly troublesome because they [GoDaddy] are such a big part of DNS.”
“We have customers who say the quality of GoDaddy data is just degrading across the board, either through direct look-ups or in some of the DomainTools products themselves,” he said.
DomainTools customers include security professionals trying to hunt down the source of attacks and intellectual property interests trying to locate pirates and cybersquatters.
GoDaddy today confirmed to DI that it has been throttling DomainTools’ Whois access, and said that it’s part of ongoing anti-spam measures.
In recent years there’s been an increase in the amount of spam — usually related to web design, hosting, and SEO — sent to recent domain registrants using email addresses harvested from new Whois records.
GoDaddy, as the market-share leader in retail domain sales, takes a tonne of flak from customers who, unaware of standard Whois practice, think the company is selling their personal information to spammers.
This kind of Twitter exchange is fairly common on GoDaddy’s feed:

Being bombarded by web developers after purchasing domain frm @GoDaddy
I paid for that domain and u selling my personal info like anything.gotta switch frm godaddy.

— Vikas Rawat (@VikasRa87555925) January 12, 2018

While GoDaddy is not saying that DomainTools is directly responsible for this kind of activity, throttling its port 43 traffic is one way the company is trying to counter the problem, VP of policy James Bladel told DI tonight.
“Companies like [DomainTools] present a challenge,” he said. “While we may know these folks, we don’t know who their customers are.”
But that’s just a part of the issue. GoDaddy was also concerned about the amount of resources DomainTools was consuming, and its own future legal responsibilities under the European Union’s forthcoming General Data Protection Regulation.
“When [Chen] says they’re down to a fraction or a percentage of what they had previously, well what they had previously was they were updating and archiving Whois almost in real time,” Bladel said. “And that’s not going to fly.”
“That is not only, we feel, not congruent with our responsibilities to our customers’ data, but it’s also, later on down the road, exactly the kind of thing that GDPR and other regulations are designed to stop,” he said.
GDPR is the EU law that, when it fully kicks in in May, gives European citizens much more rights over the sharing and processing of their private data.
Bladel added that DomainTools is still getting more Whois access than other parties using port 43.
“They have a level of access that is much, much higher than what they would normally have as a registrar,” he said, “but much lower than I think they want, because they want to effectively download and keep current the entirety of the Whois database.”
I’m not getting a sense from GoDaddy that it’s likely to backtrack on its changes.
Indeed, the company also today announced that it from January 25 it will start to “mask” key elements of Whois records when queried over port 43.
GoDaddy told high-value customers such as domainers today that port 43 queries will no longer return the registrant’s first name, last name, email address or phone number.
Bulk Whois users such as registrars (and, I assume, DomainTools) that have been white-listed via the “GoDaddy Port43 Process” will continue to receive full records.
Its web-based Whois, which includes a CAPTCHA gateway to prevent scraping, will continue to function as normal.
Bladel said that these changes are NOT related to GDPR, nor to the fact that ICANN said a couple months back that it would not enforce compliance with Whois provisions of the Registrar Accreditation Agreement, subject to certain conditions.

9 Comments Tagged: DomainTools, gdpr, godaddy, ICANN, port 43, privacy, security, throttling, whois

Big changes at DomainTools as privacy law looms

Regular users of DomainTools should expect significant changes to their service, possibly unwelcome, as the impact of incoming European Union privacy law begins to be felt.
Professional users such as domain investors are most likely to be impacted by the changes.
The company hopes to announce how its services will be rejiggered to comply with the General Data Protection Regulation in the next few weeks, probably in February, but CEO Tim Chen spoke to DI yesterday in general terms about the law’s possible impact.
“There will be changes to the levels of service we offer currently, especially to any users of DomainTools that are not enterprises,” Chen said.
GDPR governs how personal data on EU citizens is captured, shared and processed. It deals with issues such as customer consent, the length of time such data may be stored, and the purposes for which it may be processed.
Given that DomainTools’ entire business model is based on capturing domain registrants’ contact information without their explicit consent, then storing, processing and sharing that data indefinitely, it doesn’t take a genius to work out that the new law represents a possibly existential threat.
But while Chen says he’s “very concerned” about GDPR, he expects the use cases of his enterprise customers to be protected.
DomainTools no longer considers itself a Whois company, Chen said, it’s a security services company now. Only about 20% of its revenue now comes from the $99-a-month customers who pay to access services such as reverse Whois and historical Whois queries.
The rest comes from the 500-odd enterprise customers it has, which use the company’s data for purposes such as tracking down network abuse and intellectual property theft.
DomainTools is very much aligned here with the governments and IP lawyers that are pressing ICANN and European data protection authorities to come up with a way Whois data can still be made available for these “legitimate purposes”.
“We’re very focused on our most-important goal of making sure the cyber security and network security use cases for Whois data are represented in the final discussions on how this legislation is really going to land,” he said.
“There needs to be some level of access that is retained for uses that are very consistent with protecting the very constituents that this legislation is trying to protect from a privacy perspective,” he said.
The two big issues pressing on Chen’s mind from a GDPR perspective are the ability of the company to continue to aggregate Whois records from hundreds of TLDs and thousands of registrars, and its ability to continue to provide historical, archived Whois records — the company’s most-popular product after vanilla Whois..
These are both critical for customers responding to security issues or trying to hunt down serial cybersquatters and copyright infringers, Chen said.
“[Customers are] very concerned, because their ability to use this data as part of their incident response is critical, and the removal of the data from that process really does injure their ability to do their jobs,” he said.
How far these use cases will be protected under GDPR is still an open question, one largely to be determined by European DPAs, and DomainTools, like ICANN the rest of the domain industry, is still largely in discussion mode.
“Part of what we need to help DPAs understand is: how long is long enough?” Chen said. “Answering how long this data can be archived is very important.”
ICANN was recently advised by its lawyers to take its case for maintaining Whois in as recognizable form as possible to the DPAs and other European privacy bodies.
And governments, via the Governmental Advisory Committee, recently urged ICANN to continue to permit Whois access for “legitimate purposes”.
DomainTools is in a different position to most of the rest of the industry. In terms of its core service, it’s not a contracted party with ICANN, so perhaps will have to rely on hoping whatever the registries and registrars work out will also apply to its own offerings.
It’s also different in that it has no direct customer relationship with the registrants whose data it processes, nor does it have a contractual relationship with the companies that do have these customer relationships.
This could make the issue of consent — the right of registrant to have a say in how their data is processed and when it is deleted — tricky.
“We’re not in a position to get consent from domain owners to do what we do,” Chen said. “I think where we need to be more thoughtful is whether DomainTools needs to have a process where people can opt out of having their data processed.”
“When I think about consent, it’s not on the way in, because we just don’t have a way to do that, it’s allowing a way out… a mechanism where people can object to their data being processed,” he said.
How DomainTools’ non-enterprise customers and users will be affected should become clear when the company outlines its plans in the coming weeks.
But Chen suggested that most casual users should not see too much impact.
“The ability of anyone who has an interest in using Whois data, who needs it every now and then, for looking up a Whois record of a domain because they want to buy it as a domain investor for example, that should still be very possible after GDPR,” he said.
“I don’t think GDPR is aimed at individual, one-at-a-time use cases for data, I think it’s aimed at scalable abuse of the data for bad purposes,” he said.
“If you’re running a business in domain names and you need to get Whois at significant scale, and you need to evaluate that many domains for some reason, that’s where the impact may be,” he said.
Disclosure: I share a complimentary DomainTools account with several other domain industry bloggers.

5 Comments Tagged: .eu, copyright, DomainTools, gdpr, ICANN, privacy, security, trademarks, whois

Bezsonoff replaces Kaine at Neustar

.CO Internet alum Nicolai Bezsonoff has replaced Sean Kaine as head of Neustar’s domain name business.
Neustar today announced that Bezsonoff has been appointed VP and general manager of the Registry Solutions business.
That’s Kaine’s old job. I hear he’s leaving the company of his own volition, but I don’t know where he’s going.
Bezsonoff was in a similar role in the Security Solutions division.
He joined Neustar when it acquired Colombian ccTLD registry .CO, where he was COO and co-founder, for $109 million almost four years ago.
The announcement comes just a few weeks after it was announced that Afilias is to take over the running of Australia’s 3.1 million-name ccTLD .au, one of Neustar’s marquee tenants.

Comment Tagged: neustar

Active new gTLD domains drop below 20 million

The number of domain names recorded in new gTLD zone files has dipped below 20 million for the first time in 18 months.
The total crossed the milestone in the wrong direction January 1, according to DI’s records.
As of today, there are 19.8 million domains in zone files, down from a peak of 26 million in March 2017.
The count has gone down by about half a million names in the last 90 days, largely as a result of declines in .top, .xyz and .kiwi, which have each recorded six-figure losses.
It’s the first time that the zone files have showed the number of domains going below 20 million since the beginning of June 2016, when XYZ.com sold millions of .xyz domains for a penny each. Most of those names did not renew a year later.
Zone files do not record every domain that has been registered, just those with active name servers. Others may be registered but unused or on hold for various reasons.

1 Comment Tagged: new gTLDs