Bernie Sanders’ proposal for public equity stakes in AI companies faces scrutiny over effectiveness and potential conflicts of interest.
Summary
- Sanders proposes creating a sovereign wealth fund by taking 50% stakes in major AI companies like OpenAI and Anthropic
- Critics argue public ownership would create conflicts between corporate profits and public interest, citing Norwegian oil fund precedent
- Alternative approaches include AI-specific taxation and government-operated ‘public option’ AI models
- Switzerland’s Apertus project demonstrates publicly-developed AI model focusing on transparency and compliance
Sanders’ Sovereign Wealth Fund Proposal
Senator Bernie Sanders has proposed creating a US sovereign wealth fund by acquiring 50% stakes in major AI companies including Anthropic, OpenAI and xAI. The plan aims to establish democratic control over AI development through voting shares and board representation, while ensuring public benefit from AI’s economic returns. Sanders frames this as addressing the concentration of power among what he calls 86 AI billionaires who are ‘seeking to maximize their power and profit’ while determining ‘the fate of humanity behind closed doors in Silicon Valley’.
Critics Highlight Conflict Risks
Security researchers warn that public ownership could create perverse incentives, making governments stakeholders in corporate profits rather than public interest. They argue this structure would encourage regulators to clear obstacles, permit worker and user exploitation, and suppress competition to boost stock valuations. The Norwegian sovereign wealth fund’s substantial oil company holdings serves as a cautionary example, where government dependence on those investments has reportedly inhibited climate action rather than steering companies toward environmental responsibility.
Alternative Approaches Emerge
Critics propose separating wealth redistribution from corporate governance through targeted taxation rather than equity stakes. Senator Elizabeth Warren’s datacenter energy tax proposal represents one approach, while others suggest AI token taxes to capture economic benefits. For governance influence, researchers advocate for government-operated ‘AI public options’ – publicly developed models that would compete with private offerings while maintaining democratic control.
Swiss Model Shows Public AI Path
Switzerland’s Apertus project demonstrates practical implementation of public AI development. Built by public servants and university researchers using licensed training data and existing public supercomputing infrastructure, Apertus prioritises transparency, sustainability, and regulatory compliance over performance benchmarks. While not matching private models on technical metrics, it establishes competitive pressure for responsible corporate behaviour without direct ownership stakes.
Why it matters
CISOs must understand how government AI policies could reshape the vendor landscape, compliance requirements, and competitive dynamics. Whether through ownership stakes, taxation, or public alternatives, policy interventions will affect procurement decisions, regulatory frameworks, and the balance between public oversight and private innovation in AI technologies your organisation depends on.
What to do now
- Monitor development of AI taxation proposals that could affect vendor pricing and business models
- Evaluate public AI options like Switzerland’s Apertus for compliance-focused use cases
- Assess how government ownership stakes in AI companies might influence vendor relationships and regulatory requirements
- Consider transparency and compliance advantages of publicly-developed AI models for risk-sensitive applications
