As AI agents become increasingly sophisticated and indistinguishable from human users, the focus shifts from their capabilities to the trust architecture underpinning them. The challenge lies in verifying human presence and accountability in digital interactions. World ID, a proof-of-human system, exemplifies this shift by enabling users to prove their uniqueness without revealing personal details, with applications spanning dating, gaming, and online communities. Ajay Patel of World ID emphasizes that trust can no longer be inferred from behavior alone; it must be explicitly proven. This need for verifiability extends to AI agents, where increased capability does not inherently equate to increased trust. AI researcher Edoardo Contente highlights the importance of governance tools like offline evaluation and after-the-fact auditing to manage risks associated with AI agents. As the internet evolves, distinguishing between human and AI outputs becomes crucial, with accountability and verifiability emerging as key components of trust in the digital age.