Agent-to-Agent (A2A)
A protocol for interoperability between autonomous agents built by different vendors on different frameworks — capability discovery, task delegation, and structured message exchange across organisational boundaries.
Why it matters for autonomous finance
A2A is where the systemic question becomes concrete. The moment agents delegate to agents across firm boundaries, 'who decided' stops being answerable inside any single institution. Anyone writing an oversight policy that assumes decisions originate in-house should read this protocol as a description of the world their policy is about to stop describing.
What it does not cover
Interoperability protocols distribute capability faster than they distribute accountability. A2A defines how agents talk; it does not define whose balance sheet absorbs the outcome when a delegated task goes wrong, and no financial regulator has said either.
No provisions for auditability of cross-agent decisions or liability allocation. Every entry in this register carries this section. A standard read past its scope is worse than no standard, because it produces confidence without coverage.