Map the corporate supply chain behind state AI-driven harm

Identify the specific companies, contracts, data brokers, and cloud providers that make a coercive state programme — deportation, mass surveillance, predictive policing, targeted spyware — technically possible, then publish the map. The deliverable is a sourced public document naming names: contract numbers, dollar amounts, the vendor's stated role, the technical capabilities it provides. The document is then handed to investors, employees, students, and shareholders of the named vendors.

An actor chooses this strategy because the state programme is too well-defended to attack directly and too distributed across statute and discretion to legislate against. The corporate vendors are softer targets — they answer to investors who care about reputational risk, to employees who can refuse to work on a project, to universities that can drop them from research partnerships. Naming the vendor converts a diffuse political fight into a specific institutional one with clear pressure points.

It trades off attribution risk and a treadmill problem: vendors can be replaced, contracts re-routed, and the campaign has to remap on every rotation. The strategy works best when paired with a constituency (immigrant rights, civil rights, tech workers) whose pressure on the vendor is credible; without that, the map is journalism, not leverage.