Vint Cerf, one of the architects of the internet's foundational protocols, is now advising Innovation Labs on developing open standards for AI agents. Cerf recently left Google after 20 years and is now working with the organization, which aims to create an open architecture for AI agents to identify themselves. Innovation Labs, a subsidiary of Identity Digital, proposes DNSid as a registry for agent identification that links each AI agent to an existing internet domain name and uses cryptographic proofs to log its registration over time.
Cerf joined a group of internet luminaries supporting the initiative, which seeks to address the lack of a shared standard for identifying and auditing AI agents. Currently, most AI agents operate within proprietary systems, but businesses are envisioning a future where they interact directly with other agents across the internet. Innovation Labs’ interim CEO, Allie Kline, said the company is trialing the standards with unnamed hyperscalers and identity companies. Cerf emphasized the importance of establishing trust in AI agents, noting the challenges in determining accountability and identity for these entities.
Cerf highlighted that the key to widespread adoption of any protocol will be its functionality, citing the historical adoption of TCP/IP as a precedent. He also noted that Innovation Labs’ proposal does not include broader plans for AI business or ownership of registration data. Kline added that there is skepticism toward hyperscalers releasing standards due to concerns about proprietary data. Cerf remains cautious about the inevitability of an agentic economy but acknowledges the likelihood of people adopting AI agents for convenience.
Source: techcrunch
