Artificial intelligence is changing the way we use the internet. We're moving beyond asking AI questions and receiving answers. Increasingly, AI agents will be able to make bookings, purchase goods, manage accounts and communicate with other systems on our behalf.
That brings a whole new set of responsibilities.
If an AI agent contacts your business, how do you know who it represents? How can you be sure it has permission to act? And who is responsible if something goes wrong?
Singapore is looking at how the internet's existing naming infrastructure might help answer those questions.
Singapore Turns to DNS
On 4 November 2026, Singapore's Parliament discussed the possibility of creating a national digital identity registry for AI agents.
Member of Parliament Tin Pei Ling proposed a system that would allow organisations to identify AI agents and verify who was responsible for them.
In response, Senior Minister of State Jasmin Lau confirmed that Singapore's Infocomm Media Development Authority (IMDA) is already working with GoDaddy, Infoblox, Cloudflare and Google to explore how the Domain Name System, or DNS, could help identify and verify AI agents.
Singapore has not decided to establish a national registry. The government is first examining whether existing internet technologies and international standards can do the job.
Meanwhile, GovTech is developing an internal registry to keep track of AI agents used by government agencies and the people responsible for them.
The discussion has moved beyond giving AI agents recognisable names. Governments are beginning to consider how those names connect to identity, responsibility and trust.
The Technology Is Already Taking Shape
The Singapore announcement follows a demonstration on 14 October 2026 involving GoDaddy, Infoblox, Google and Cloudflare.
Working with IMDA, the companies demonstrated a system called Chain of Trust, showing how AI agents operating on different platforms could discover one another, verify their identities and complete a transaction.
The example involved an AI agent purchasing a flight from Madrid to Singapore.
Rather than requiring both agents to operate within the same company's platform, the demonstration used existing internet standards to help them communicate securely.
DNS played a role in helping agents discover and verify one another, while other technologies handled authorisation and the transaction itself.
It is an encouraging direction. The internet has always benefited from systems that allow different networks, services and organisations to communicate.
A Name Is More Than an Address
Most of us rarely think about the systems behind a domain name. We type an address into a browser and expect to arrive at the right destination.
Over time, those names have also become part of our business identities.
A familiar domain tells customers where to find us. It appears on our websites, email addresses, advertising and business documents.
AI agents could eventually become another part of that identity.
Imagine a customer dealing with an AI agent that appears to represent a business they already know and trust.
The name might look familiar, but that alone doesn't prove the agent is genuine. There still needs to be a reliable way to verify who operates it and what it is authorised to do.
Singapore's work is exploring exactly those kinds of safeguards.
Yet there is another part of the conversation that shouldn't be overlooked.
Who controls the name?
Where the Handshake Protocol Comes Into the Picture
Handshake offers a different approach to internet naming.
Instead of relying entirely on the conventional DNS root system, Handshake allows people and organisations to control top level names through a decentralised blockchain protocol.
For businesses looking towards a future involving AI agents, that distinction could become increasingly relevant.
A business might want its agents to operate under names that reflect its established brand, without having to depend entirely on another company controlling the naming infrastructure.
There is an important difference, however, between owning a Handshake top level name and registering a name beneath a top level domain controlled by someone else.
egistering yourbrand.agent does not necessarily mean you own the .agent extension itself.
Handshake also does not automatically prove that an AI agent is genuine, authorised or trustworthy. Those protections still require additional systems.
And because Handshake operates outside the conventional DNS root, wider compatibility and adoption remain practical challenges.
What Handshake offers is another way to consider who controls the underlying name.
As the internet moves towards greater use of AI agents, that is a discussion worth having.
Who Controls the Identity?
There is a lot to welcome in Singapore's approach.
AI agents will need reliable ways to identify one another, establish trust and operate safely across different platforms.
The involvement of major DNS companies also shows that existing internet infrastructure could play an important role in making this possible.
Open standards are particularly encouraging because they can reduce the need for everyone to depend on a single commercial platform.
But being able to communicate across platforms isn't necessarily the same as controlling your own name.
A business might have an AI agent that works with hundreds of different services while still depending on a third party to maintain the name under which that agent operates.
That may be perfectly acceptable for many businesses. Others may prefer greater independence, particularly when their names and digital identities become valuable parts of their operations.
Neither approach removes the need for security, accountability or proper identity verification.
The difference is how much control the name holder retains.
A Conversation Worth Having Now
Singapore hasn't settled on a national registry, and the standards surrounding AI agent identity are still developing.
That makes this an important time to consider the choices being made.
We have already seen how much value businesses place on domain names and online identities. AI agents are likely to add another dimension to that relationship.
As these technologies become part of everyday business, the names attached to them may become just as important as the services they provide.
Singapore's announcement is a reminder that the next stage of internet naming is already being discussed.
And while much of the attention will understandably focus on making AI agents secure and trustworthy, we should also be thinking about the independence of the people and businesses they represent.
After all, if an AI agent is going to represent your business, shouldn't you have a meaningful say in who controls its name?
Explore Handshake Naming
If you're interested in giving your AI agent a recognisable name using Handshake, HeadlessDomains.com offers a way to explore and register names beneath Handshake extensions such as .agent.
These names operate outside the conventional ICANN DNS root. Registering a name beneath an existing extension is different from owning the extension itself, and a name alone does not verify an agent's identity or authority.
Sources
- Channel NewsAsia, Singapore parliamentary debate, 4 November 2026
- AsiaOne, Singapore's AI agent safeguards, 4 November 2026
- GoDaddy, Chain of Trust demonstration, 14 October 2026
- The Straits Times, Singapore's government AI registry, 2 June 2026
Own the Name | Own the Rules | Own The Dot.