About

A public evidence resource for AI safety and governance.

What This Site Is

AI Democratic Governance is a public evidence resource tracking AI capabilities and the real-world incidents they manifest. It is UK-focused with global context.

The capabilities catalogue is the core of the site — sourced entries on what AI systems can demonstrably do, classified by where the capability lives (in the raw model, with harness, or both) and by evidence stage (open source, frontier-model, or hypothesised). Every entry carries a sourced counterargument. The methodology is documented in full: source-quality hierarchy, inclusion criteria, and the editorial process by which classifications are made and revised.

Separately, the Advocacy section offers practical tools for democratic engagement on AI governance — template letters to MPs and guidance for following Parliamentary proceedings. It is kept clearly separate from the evidence layer.

Why It Exists

The people building AI are warning about its risks, but the information rarely reaches ordinary readers in a usable form. Academic papers sit behind jargon. Policy analysis is scattered. Tools for civic engagement are not gathered in one place.

This site closes that gap. It is built to be the kind of resource that journalists can cite, Select Committees can reference, and citizens can use to hold their representatives to account.

Principles

What This Site Is Not

Contact

A general contact mechanism is forthcoming. In the meantime, errors, broken source links, and missing counterarguments can be reported via the site's GitHub repository.

Get Involved

Explore the evidence, then take action — write to your MP, follow Parliamentary proceedings, or share what you learn with someone who needs it.