What the Trust Layer Can Currently Deliver

The trust layer is best understood as a set of overlapping but distinct functions, each with different technical feasibility profiles and different relationships to the constitutional constraints.

The most technically mature function is verified attestation: the ability to record specific claims about a participant's characteristics or history — that they have completed a particular governance process, that they have been attested to by specific other participants, that they have maintained a particular commitment over a specified period — in a way that is verifiable by others without being falsifiable by the subject and without revealing information beyond what the specific attestation claims. Verifiable credential technologies, combined with zero-knowledge proofs, can currently deliver this function for specific, bounded claims with reasonable privacy preservation.

The less technically mature function is contextual reputation: the ability to communicate the quality of a participant's relational contribution across contexts in a way that is meaningful to recipients while being resistant to gaming and respectful of the participant's sovereignty over their own record. This function is harder because relational contribution is contextual in ways that attestation is not: what makes someone trustworthy in one relational context may not make them trustworthy in another, and a system that aggregates reputation across contexts risks obscuring the contextual nature of trust in ways that are both epistemically distorting and constitutionally problematic.

The minimum viable trust layer — the version that can be built while the more difficult functions continue to be developed — focuses on verified attestation for specific, consensual, bounded claims, with explicit acknowledgment that the more ambitious functions of contextual reputation require further technical and governance development before they can be implemented in ways compatible with the constitutional principles.

The social science literature on trust and trustworthiness is relevant to the trust layer's design in ways that the technical literature sometimes obscures. Trust, as analyzed by scholars from Diego Gambetta's edited volume Trust: Making and Breaking Cooperative Relations (1988) through Onora O'Neill's work on trust and transparency (A Question of Trust, 2002) to recent work by Rachel Botsman on distributed trust, is not a property of individuals that can be measured and aggregated. It is a relational phenomenon that emerges in specific contexts through specific experiences. The trust layer must be designed with this understanding of trust rather than with the simplified model — trust as a score — that makes the technical problem more tractable but the result constitutionally problematic.

The mechanism design literature on reputation systems provides the technical foundation for understanding why trust scores are vulnerable to gaming and what design choices affect their gaming resistance. Paul Resnick and Richard Zeckhauser's foundational analysis of eBay's reputation system, the extensive subsequent literature on reputation systems in online markets, and the more recent work on adversarial robustness in machine learning systems all contribute to understanding the technical side of the gaming problem. The constitutional response to gaming vulnerability — treating the gaming of trust signals as a constitutional violation rather than merely an operational problem — is a governance response that must accompany the technical design.

The deepest open question about the trust layer concerns the relationship between trust and power. In every social system in which trust is a significant resource, the distribution of trust is correlated with prior distributions of power. People who have access to well-established networks, prestigious institutional affiliations, and documented histories of consequential participation will generate trust signals more easily than people who lack those advantages — not necessarily because they are more trustworthy, but because their trustworthiness has been more legibly documented in forms that the trust layer can recognize. A trust layer that reproduces existing power distributions under the label of merit is not a constitutional innovation. It is an existing hierarchy with a new verification mechanism. How the trust layer is designed to resist this reproduction — while still performing the function of distinguishing genuine trustworthiness from its simulation — is a question for which the intellectual traditions engaged in this chapter do not provide a fully satisfying answer.