Latest news of the domain name industry

Recent Posts

Melbourne IT scales back HARM proposal

Kevin Murphy, November 14, 2012, Domain Policy

Melbourne IT has published a revised, less-complicated version of its High At-Risk Marks (HARM) proposal for protecting famous brands in the new gTLD program.

The new version throws more than a few bones to trademark lawyers, most of whom rejected many aspects of the original proposal at a meeting in Washington DC this September.

It’s a lot closer to the eight-point wish-list published jointly by the Intellectual Property Constituency and Business Constituency last month.

HARM envisions a two-tier set of trademark rights protection mechanisms in new gTLDs, with the super-famous brands that get cybersquatted and phished on a regular basis enjoying greater privileges.

Companies that could prove their trademarks were subject to regular abuse would, for example, benefit from a perpetual Trademark Claims notification service on “brand+keyword” domains.

The new version would lower the bar for inclusion on the list.

The first HARM said trademarks should be registered on five continents, but the new version reduces that to a single registration, provided that the jurisdiction does substantive review.

A provision to only extend the protection to five-year-old marks has also been removed, and the number of UDRP wins required to prove abuse has also been reduced from five to one.

I’ve previously expressed my fondness for the idea of using UDRP decisions to gauge the risk profile of a trademark, but it was recently pointed out to me that it may incentivize mark holders to pay people to cybersquat their marks, in order to win slam-dunk UDRPs and thus benefit from better RPMs, which makes me less fond of it.

Even if such skullduggery is an outside risk, I think a single UDRP win may be too low a bar, given the number of dubious decisions produced by panelists in the past.

The revised HARM would still exclude dictionary words from the special protections (as the paper points out, Apple and Gap would not be covered). The proposal states:

Melbourne IT believes it will be difficult to get consensus in the ICANN community that this mechanism should apply to all trademark owners, most of whom do not suffer any trademark abuse. Many trademarks also relate to generic dictionary words that would be inappropriate to block across all gTLDs.

The original HARM paper was put forth as compromise, designed to help prevent or mitigate the effects of most cybersquatting, while being slightly more palatable to registries and registrars than the usual all-or-nothing demands coming from trademark lawyers.

While not particularly elegant, most of its recommendations were found wanting by the ICANN community, which is as bitterly divided as always on the need for stronger rights protection mechanisms.

The IPC and BC did adopt some of its ideas in their recent joint statement on enhanced RPMs, including the idea that frequently squatted names should get better protection, but rejected many more of the Melbourne-proposed criteria for inclusion on the list.

Meanwhile, many registrars shook their heads, muttering something about cost, and new gTLD applicants staunchly rejected the ideas, based on the mistaken notion that paying their $185,000 has rendered the Applicant Guidebook immutable.

Read the new Melbourne IT paper here (pdf).

What’s wrong with Melbourne IT’s new anti-cybersquatting plan?

Kevin Murphy, August 16, 2012, Domain Policy

Genuine question.

Melbourne IT, the Aussie registrar with the increasingly vocal brand-protection focus, has come up with a new scheme for protecting super-famous brands after new gTLDs start to launch.

It draws on elements of the abandoned Globally Protected Marks List, ICM Registry’s Sunrise B policy, .CO Internet’s launch program, and various recent demands from the intellectual property community.

It’s called the paper Minimizing HARM (pdf), where HARM stands for High At-Risk Marks.

The title may set off grammatical alarm bells, but the rest reads like the least-unreasonable proposition for protecting big brands from cybersquatters that I’ve come across in a long time.

What I like about it is that it’s actually contemplating ways to prevent gaming from the outset, which is something the IP lobby hardly ever seems to do when it demands stronger rights protection mechanisms.

The idea calls for the forthcoming Trademark Clearinghouse to flag a narrow subset of the trademarks in its database as High At-Risk Marks that deserve special treatment.

Melbourne IT has organizations such as PayPal and the Red Cross in mind, but getting on the list would not be easy, even for famous brands.

First, companies would have to prove they’ve had trademark protection for the brand in three of ICANN’s five geographic regions for at least five years — already quite a high bar.

Implemented today, that provision could well rule out brands such as Twitter, which is an obvious high-risk cybersquatting target but might be too young to meet the criteria.

Dictionary words found in any of UN’s six official languages would also be banned, regardless of how famous the brand is. As the paper notes, that would be bad news for Apple and Gap.

Companies would also have to show that their marks are particularly at risk from phishing and cybersquatting.

Five successful UDRP complaints or suspensions of infringing domains by a “top ten registrar” would be enough to demonstrate this risk.

But that’s not all. The paper adds:

In addition to meeting the minimum criteria above, the High At-Risk Mark will need to obtain a minimum total points score of 100, where one point is awarded for each legal protection in a jurisdiction, and one point is awarded for each successful UDRP, court action, or domain registrar suspension undertaken in relation to the mark.

That appears to be setting the bar for inclusion high enough that an OlympicTM pole-vaulter would have difficulty.

Once a brand made it onto the HARM list, it would receive special protections not available to other brands.

It would qualify for a “Once-off Registration Fee”, pretty much the same as ICM’s .xxx Sunrise B, where you pay once to block your exact-match domain and don’t get pinged for renewal fees every year.

Any third parties attempting to register an available exact-match would also have to have two forms of contact information verified by the gTLD registry before their names resolved.

The Trademark Claims service – which alerts mark owners when somebody registers one of their brands – would run forever for HARM-listed trademarks, rather than just for the first 60 days after a gTLD goes into general availability.

The always controversial Uniform Rapid Suspension service would also get tweaked for HARM trademarks.

Unless the alleged cybersquatter paid the equivalent of a URS filing fee (to be refunded if they prevail) their domains would get suspended 48 hours after the complaint was filed.

I’m quite fond of some of the ideas in this paper.

If ICANN is to ever adopt a specially protected marks list, which it has so far resisted, the idea of using favorable UDRP decisions as a benchmark for inclusion – which I believe Marque also suggested to ICANN back in February – is attractive to me.

Sure, there are plenty of dumb UDRP decisions, but the vast majority are sensible. Requiring a sufficiently high number of UDRP wins – perhaps with an extra requirement for different panelists in each case – seems like a neat way of weeding out trademark gamers.

The major problem with Melbourne IT’s paper appears to be that the system it proposes is just so complicated, and would protect so few companies, that I’m not sure it would be very easy to find consensus around it in the ICANN community.

I can imagine some registries and registrars might not be too enthusiastic when they figure out that some of the proposals could add cost and friction to the sales process.

Some IP owners might also sniff at the some of the ideas, just as soon as they realize their own trademarks wouldn’t meet the high criteria for inclusion on the HARM list.

Is Melbourne IT’s proposal just too damn sensible to pass through ICANN? Or is it riddled with obvious holes that I’ve somehow manged to miss?

Discuss.