Beyond AI Philanthropy: Measuring the Distribution of AI-Generated Wealth

This research note examines the proposed Anthropic Public Benefit Fund as an emerging example of how artificial intelligence companies may allocate a portion of their economic value toward public benefit. The analysis distinguishes voluntary corporate contributions from the broader distribution of AI-generated economic gains. It proposes a preliminary framework for evaluating four dimensions: economic value creation, distribution of economic benefits, measurable socioeconomic outcomes, and governance and accountability. Using Anthropic's public announcement as an initial case, the note identifies five research questions and three hypotheses concerning economic security, beneficiary participation, and the measurement of public benefit. It also discusses limitations relating to available evidence, causal attribution, and the absence of verified outcomes from the proposed fund. The central argument is that commitments to public benefit should be assessed not only by the resources pledged, but also by their governance, distribution, and demonstrable effects on people's economic circumstances. This is an independent, preliminary research note prepared within the RISEN Research Initiative. It is not a peer-reviewed study, and its proposed framework has not yet been empirically validated.