It’s highly unlikely registrants will lose their domains because of the UK’s treaty.
You’ve likely seen several articles over the past week warning that .io domain names might be in jeopardy.
Years ago, I would have jumped on this bandwagon. But I don’t think .io domain owners have anything to worry about.
What’s happening?
Like all two letter top level domains, .io is a country code top level domain name (ccTLD). It’s the two-digit code for British Indian Ocean Territory, an overseas territory of the United Kingdom located in the Indian Ocean.
After much pressure, the UK has agreed to a deal that will hand over control of the territory to the country of Mauritius.
This means that the British Indian Ocean Territory might cease to exist. And that technically means trouble for .io domains.
ICANN defers to ISO, a standards organization, when deciding which country code domains should exist and what their two-letter code is. If ISO boots IO, a process would begin to retire the top level domain.
In 2008, I asked ICANN what would happen if the island nation of Tuvalu ceased to exist as rising waters swamped the country. Its .tv country code has become popular with video sites. ICANN told me:
The simple answer is if a country ceases to exist, under existing policy the domain will need to be retired. ICANN will permit a reasonable transition period so the affected parties can transition to other domains. In the most recent case, .YU (replaced with .RS and .ME), there was a 3 year transition period assigned by the ICANN Board.
So people are right to question .io’s future. .Yu isn’t the only ccTLD to be removed, although not all domains have followed this process.
And yet, there’s a pragmatic side to this issue. ICANN’s goal is the security and stability of the domain name system. What impact does yanking a popular top level domain, used by small and large companies alike, have on the domain name system? It would be counter to ICANN’s mission.
More likely scenarios
While I don’t think ICANN would move to delete the .io namespace, there are some practical implications.
The people sweating the most right now are at Identity Digital, the company that runs the .io namespace.
Identity Digital acquired the namespace when it bought Afilias a few years ago. .Io was likely a key value driver for the acquisition.
A couple of things could happen to Identity Digital’s deal for .io.
One is that nothing happens. ICANN allows it to continue running .io, and the name isn’t “handed over” to Mauritius.
The other is that the domain is handed to Mauritius, or the country negotiates a deal with Identity Digital to take a cut of revenue.
The latter could impact domain registrants because they might end up paying more for registrations and renewals in the future.
Bigger risks exist for other ccTLD users
While I don’t think .io registrants have much to worry about, I believe it’s important for ccTLD registrants to understand more about the domains they are using.
ccTLDs are the wild west of the internet. ICANN has little oversight over these domains, and each country can set its own policies.
Was it wise for Notion to set up shop on .so? That’s Somalia’s domain name, and that’s not the most stable country. (The company later bought notion.com).
And how about when bit.ly started on .ly? That’s Libya’s domain and comes with many risks.
Unstable countries are only part of the problem.
A hot ccTLD right now is .ai, which many artificial intelligence companies have adopted.
.Ai is the country code for the island of Anguilla and its 15,000 inhabitants.
Over 500,000 .ai domains are now registered. But basically, one guy in Anguilla sets the policies for the domain, and those policies have changed.
There are many unusual rules for using .ai domains that can change on a whim.
(That’s not to mention what a Hurricane can do. While .ai has become more robust over the years, I sometimes hear about similar issues in other ccTLDs.)
The takeaway
Domain registrants should be aware of what they are registering when they choose a two letter domain. Many of these country code domains have little oversight, which introduces a risk to anyone using them. However, I believe the current uproar over .io is overblown.






Andrew makes a strong argument.
But I disagree with his conclusion.
Everything swirling around the (.io) power struggle is soaked in a debate over colonialism.
I’m betting (.io) is retired to bury that British history for Mauritius and the Chagossians.
If I was a founder or VC board member of a startup sitting on (.io), I’d be very worried.
Why would Western startups want to be culpable for funding digital colonialism? There is the dispute between UK & Mauritius/Changossians. But there is another dispute between Mauritius and the Chagossians.
Can you imagine what the matching dotcom pricing will be the day after ICANN/IANA announce the retirement of (.io)?
I would expect forward thinking founders and Boards to advise their teams to do everything possible to secure the dotcom within the next year or so.
If Andrew is wrong he won’t be required to subsidize any startups who lose betting on (.io).
There are possibilities other than removing the TLD, however. If the TLD is kept alive, but re-delegated, then the new TLD manager can set new rules and/or refuse to renew existing domains.
For example, .la is the ccTLD for Laos. Quite a number of years ago, that TLD was re-packaged and re-launched in the hopes of attracting registrants targeting Los Angeles. In order to do that, they dumped a lot of the existing registrants at the time.
Would that be a good move for the eventual .io manager? Probably not, and some kind of plan to deal with existing registrants could be a consideration in a redelegation proposal. However, it can also be difficult to say that a ccTLD cannot be run with due regard to the cultural or legal norms of the relevant government which is nominally expected to have some input into the process (along with the affected community, etc.).
What is the general take of Mauritians and their government relative to some of things for which .io names are used? I don’t know the answer to that question but, again, there are potential consequences of redelegation that are distinct from the question of whether a TLD is removed entirely.
Great points as always, John
The uncertainty is big blow for .IO, that is for sure, would expect values to drop.
Mr. Allemann,
I’m very unclear on why you’re making the prediction you’re making, because I see nothing in this article other than you don’t think they will because… you don’t think they will.
Your title: “.IO domain names aren’t going away.” Subtitle: you claim it’s “highly unlikely”
“.io is a […] ccTLD” for the British Indian Ocean Territory and you say it “might cease to exist” — I’m not sure that this is correct because from everything I can understand, it will definitely cease to exist. Under what conditions would it remain?
You asked ICANN what would happen if a nation ceased to exist, and state that they replied “if a country ceases to exist, under existing policy the domain will need to be retired.”
The *only* counterargument I’m seeing is that there’s a “pragmatic side to the issue,” because it would be as you stated “counter to ICANN’s mission,” offering no specifics other than your beliefs and opinions.
The rest of this article is a collection of what you think personally would be better options (which you’ve titled “more likely scenarios” but offer no evidence as to *why* they’re more likely, especially since they run 180 degrees counter to what ICANN stated is their policy), and that there are in your opinion other ccTLD’s at more risk than .io, which is neither right nor wrong but is pure speculation and offers no germain argument to the actual matter at hand as stated in your title.
Your “takeaway” is that people need to “be aware of what they are registering when they choose a two letter domain” which is absolutely correct, and is entirely unhelpful to the argument that the .io domain isn’t “going away.”
What *evidence* do you have that this is the case? While the issue isn’t necessarily cut and dry, the only actual evidence you offered which is relevant to the argument is that ICANN states unequivocally that if the country ceases to exist the ccTLD .io *WILL* in fact go away.
Respectfully,
Josh
Andrew’s right, and the reporting as always researched and found some balance to even out the strange alarmism about the .IO somehow getting Thanos-snapped due to the transition of the islands.
Currently, .SU for the Soviet Union is on a 50 year code retirement with the ISO, who maintain the ISO-3166 that the ccTLDs are based upon.
If the .IO were to be put on a similar retirement, it might enjoy the consistency of that timeline.
Not sure why so many folks want to neg the .IO TLD as much as they are, unless maybe they are seeking to sink the prices and afford names.
Take a high-level view of this… Is anyone asking the Chagosians about if they want to kill a potential revenue stream?
Not sure what kind of exports or national revenue streams might exist in the islands.
If you look at .AI, .CO or .TV where the ccTLD are popular and there are TLD incomes, before one would start to bring their thinking in the direction of there being a desire to shutter the .IO ccTLD, first ask why there would be any attraction to distrupting potential income to the islands.
Interesting discussion. I have a somewhat different outlook though.
There are a number of fundamental differences between ccTLDs and gTLDs.
It is much more difficult for Icann to wield any influence in ccTLD matters since they are assets of sovereign nations after all.
Beyond delegation and redelegation, IANA and Icann do not have much of role here.
ICANN gets a fee on gTLD registration/renewals. They don’t make money on ccTLDs and therefore have no financial incentive.
“ICANN’s goal is the security and stability of the domain name system”
Well that’s the stated goal but they do not always act in the interest of the registrants. Think of the perpetual contract granted to Verisign. Or the wild west of the nTLD program. Price caps?
Their actions here have nothing to do with security and stability, in fact they foster risk and instability by introducing tons of TLDs that have no demand and are doomed to fail.
Yes, .io is popular but other than the registrants themselves who really cares? I don’t have stats handy (don’t think this is public information), but the number of registered domain names in .io may be less than people think. Making it a public interest issue looks challenging while dealing with so many crises on all fronts.
There is now a retirement procedure that is more formalized that in the early days and granting exceptions like .su is unlikely.
Conclusion: there is a real risk that .io will ultimately be retired, in fact there is always a risk when you go for shaky extensions. And indeed, quite a few other ccTLDs present risks, that registrants are not aware of. Always do research.
I should also add that knowing the sad story of Diego Garcia I would never have bought a .io domain anyway.