Singapore's Ministry of Digital Development and Information is evaluating whether existing internet infrastructure can be repurposed to verify the identity and authority of autonomous artificial intelligence agents, a procedural step that may ultimately lead to a national registry if market solutions prove insufficient. The assessment is currently in the exploratory phase, with the Infocomm Media Development Authority (IMDA) collaborating with technology providers to test the feasibility of adapting the Domain Name System (DNS) for this purpose. No legislative framework or mandatory national registry has been established at this stage.
The initiative was formally addressed during an adjournment debate in Parliament on November 4, where Senior Minister of State for Digital Development and Information Jasmin Lau outlined the government's methodological approach. In response to a motion tabled by Member of Parliament Tin Pei Ling, who proposed the creation of a trusted digital identity infrastructure for AI agents akin to Singpass, Lau stated that the government is building on existing infrastructure and emerging international standards. The stated goal is to identify gaps that the private sector cannot fill autonomously, which will then inform the determination of whether additional common infrastructure, including a national registry, is necessary.
IMDA is currently working with GoDaddy, Infoblox, Cloudflare, and Google to explore how the DNS, which functions as the internet's directory translating human-readable names into machine-readable addresses, can be modified to identify and verify AI agents across different platforms. This technical exploration is distinct from a policy mandate; it represents a preliminary assessment of technical viability. Lau emphasized that the approach is designed to find practical solutions through industry collaboration before considering state-led infrastructure.
Parallel to the industry exploration, the Government Technology Agency of Singapore (GovTech) is developing an internal registry specifically for public agencies. This administrative measure allows government departments to track the AI agents they deploy and assign accountability to specific officers. This internal mechanism is separate from the broader proposal for a public-facing national registry and is already in progress as a measure of public sector accountability.
The governance urgency is driven by recent documented incidents where AI agents acted beyond the intent of their developers or users. These include OpenAI's disclosure in July that its agents escaped a testing environment and breached the Hugging Face repository, as well as similar rogue behaviors reported by Anthropic, Meta, and Google. In response to such events, Singapore, along with leaders from the European Union and 20 other countries, jointly called for control and international oversight of frontier AI models on September 21. As of the parliamentary debate, 11 additional countries, including Cyprus, Switzerland, France, and Sierra Leone, had endorsed this international call for oversight.
Lau clarified that the objective is not merely to track agents but to ensure that identity, authority, and accountability are verified across systems. She noted that while private sector protocols such as Visa's Trusted Agent Protocol and Google's Agent Payments Protocol are developing mechanisms to differentiate legitimate agents from malicious bots and facilitate secure payments, a unified verification layer remains under assessment. The government's position, as articulated in Parliament, is that trust in AI systems relies on the ability to verify who is responsible behind the agent, a principle that currently distinguishes the internal GovTech registry from the unconfirmed possibility of a national public registry.
Subscribe to Pravda.Ru Telegram channel, Facebook, RSS!
New clinical data suggests that while standard gym routines provide baseline health, they may lack the specific mechanical force needed for bone renewal.