AI governance does not happen in a political vacuum. It happens in a context shaped by the interests, resources, and power of different actors, and Marxist analysis is particularly well equipped to examine how class interests shape regulatory outcomes.

The largest AI companies are among the most significant political actors in any country where they operate. They employ armies of lobbyists, fund think tanks and research institutes, place former executives in regulatory positions, and shape the terms of public debate about AI through the platforms they control. Their interest in AI regulation is clear: they want frameworks that create the appearance of accountability while preserving their freedom to operate, that impose compliance costs their smaller competitors cannot afford, and that protect their control over data and AI infrastructure.

This does not mean that all AI regulation is captured by corporate interests. Public pressure, civil society advocacy, and genuine regulatory commitment have produced meaningful interventions, the EU AI Act, FTC enforcement actions, data protection regulations. But Marxist analysis asks us to be clear-eyed about the structural pressures that shape regulation, and to evaluate regulatory outcomes against the interests of the most economically vulnerable people affected by AI, not just against the preferences of well-resourced stakeholders.

Corporate AI ethics as ideology. Many large AI companies have published AI ethics principles, established ethics boards, and made public commitments to responsible AI. Marxist analysis approaches these commitments with scepticism, not because the individuals involved are necessarily insincere, but because corporate AI ethics functions ideologically. It substitutes self-regulation for external accountability. It produces the appearance of ethical commitment while preserving the structural conditions that generate harm. When ethics boards are disbanded, as Google’s was, or when ethics commitments conflict with commercial interests, as they consistently do, the commercial interests prevail.

Labour interests in AI governance. The workers most affected by AI, those whose jobs are automated, whose conditions are intensified by algorithmic management, whose data is extracted, are rarely represented in AI governance conversations. Trade unions have begun to engage more systematically with AI issues, but their voice in regulatory processes remains weak relative to that of employers and technology companies. A Marxist perspective on AI governance insists that labour interests must be centred, not as an afterthought but as a structural requirement for governance that serves the many rather than the few.

Reflection question: Who has a seat at the table in AI governance conversations you are aware of? Who does not? What would change if labour and community interests were as well-resourced as corporate interests in those conversations?