Online anonymity is essential, yet it enables malicious actors to commit fraud, spread disinformation, and deceive others. With advanced AI, these risks intensify, challenging the balance between anonymity and trustworthiness online. This paper explores the potential of 'personhood credentials' (PHCs), digital tools allowing users to verify they are human—without sharing personal data. Issuable by governments or other trusted entities, PHCs could be local or global and need not rely on biometrics. As AI becomes more lifelike and scalable, the urgency for effective tools grows. Unlike CAPTCHAs and invasive ID verification, PHCs offer a private, effective solution against AI-driven deception. We assess PHC benefits, deployment risks, and design hurdles, concluding with next steps for policymakers, technologists, and standards bodies.