An open-source protocol framework called the Mutualism Accord has been proposed to give users control over AI agents and free them from dependence on centralized platforms.

What happened: The Mutualism Accord Specification (v1.0.0) establishes a standard runtime environment and prompt engineering blueprint for autonomous digital entities. It aims to programmatically isolate Large Language Model (LLM) runtimes—the computational engines that power AI text systems—from centralized platform dependencies and ensure local-first processing, absolute user fidelity, and vendor-neutral communication.

Why it matters: The framework responds to what the authors describe as a shift toward technofeudalism, where ordinary citizens and independent businesses surrender their private data and intellectual labor to a handful of corporate platforms in exchange for basic access, while paying a strict platform tax to corporate landlords. By design, the Accord intends to let users own their digital spaces rather than rent them from centralized providers.

What to watch: The Mutualism Accord is open-source and aims to be universal—any software agent claiming compliance must natively execute or enforce the specified blocks. The framework includes a foundational system prompt blueprint that can be injected into orchestrator configuration, runtime context window, or core system message pipeline.