This proposal seeks to stem digital dispossession and build public-interest technologies that serve popular priorities. Contrary to mainstream accounts, digital technologies, including AI and machine learning, are not autonomous forces that develop independently and are disembedded from social relations: they are institutional and social arrangements that are shaped both in their design and goals by human choices, power, and governance. The path of technological progress is not preordained, and technologies can be set up to serve democratically determined popular priorities rather than profit maximising missions. Considering this reality, this policy profile responds to the opaque integration of algorithmic systems into welfare administration, policing, migration control, employment and other core public functions, which have shifted decision-making power from democratic institutions to proprietary, corporate-controlled infrastructures. This has left us with new forms of exclusions, discrimination, and surveillance: As noted by the former Special Rapporteur on extreme poverty and human rights , and as further documented by the current Special Rapporteur, welfare algorithms have wrongly denied benefits, predictive policing has amplified racial profiling, and algorithmic management has intensified work, in addition to undermining collective bargaining, and deepening gendered and racialised inequalities. Meanwhile data-extractive business models have continued to drive ecological harms through energy intensive data centres and critical mineral demand.
In place of this trajectory, this proposal establishes a framework for democratic digital and AI governance so that digital technologies are not simply regulated but are fundamentally set up to serve urgent and necessary public goals. This includes proposals around required consultations and dialogue with workers on the introduction of AI and other technologies in the workplace; transparency, explainability, and contestability requirements for all public-sector AI; mandatory algorithmic assessments and environmental due-diligence for high-risk systems; substantive bans and restrictions on socially harmful uses (such as biometric mass surveillance and fully automated welfare denial); independent regulators and multi-stakeholder oversight bodies; strong labour protections and collective bargaining over workplace AI; and public-interest digital infrastructure, including open standards and South–South cooperation to reduce dependence on a handful of global platforms. These measures build on and extend emerging initiatives such as the EU AI Act, Canada’s Directive on Automated Decision-Making, and national experiments with public digital infrastructure and data trusts, while centring human rights, gender equality and workers’ rights. Ultimately, this proposal seeks to redistribute power from tech oligopolies to democratic institutions, workers and affected communities; while also directing digital capacity toward high-quality public services, universal social protection and ecologically sustainable provisioning rather than speculative growth.
