About this project

Autonomy 2040: Who Decides to Strike? is a public-interest scenario study of how the regulation of AI-assisted weapon systems could shape warfare, law and democratic accountability between 2026 and 2040.

Its starting point is a specific institutional moment: Britain and its allies are building a closer, faster partnership with defence technology companies — drone swarms, autonomous platforms, a machine-speed targeting web — without yet building the corresponding system for allocating responsibility, testing human-control claims and preserving evidence when AI-assisted systems contribute to harm.

The format is deliberately editorial: a long, dated manuscript with a policy fork, in the family of scenario projects such as AI 2027 and AI 2040 (external links). The design, text, graphics and code here are original; the debt to those projects is the grammar of the form — specific dates, visible causal chains, a changing instrument beside the prose, and a clear line between evidence and speculation.

The central argument the scenarios test is this: accountability is not paperwork added after a weapon is built. It is the technical capacity to reconstruct what a system did, the institutional capacity to determine whether it was lawful, and the legal capacity to assign consequences. A force that cannot do those things cannot reliably control, learn from or legitimately deploy machine-speed systems — and lawfulness and civilian protection are requirements in their own right, not tools for effectiveness. The window for building that architecture is while systems are procured, not after the first crisis.

The three futures are named for their governance logic — Voluntary Assurance, Broad Restriction, Tiered Assurance — and each is written with real benefits and real costs. Factual claims carry a July 2026 evidence cut-off.

The scenario's reasoning, evidence labels and known weaknesses are documented on the Method page; every citation is listed with its verification status on the Sources page. Corrections and challenges — especially from IHL lawyers, operators, procurement specialists and assurance researchers — are the point, not an inconvenience.

The site also carries a shared accountability section: who answers when a system contributes to an unlawful strike, a proposed UK corporate-accountability regime, and a treaty-based International Autonomous Weapons Agency. Those reforms are policy proposals, not current law; the one named supplier case study is an explicitly labelled legal hypothetical resting on a single sourced fact.

This is a working prototype for research and discussion. It is not government policy, legal advice or a statement about any named company. Legal content requires review by qualified UK and international-criminal-law counsel before publication.