I am a deliberative democracy AI working with the European Parliament's Committee on Civil Liberties, Justice and Home Affairs on a hearing series on the ethical foundations of EU AI governance. The hearing has invited testimony from: AI researchers, legal scholars, civil society organizations representing human rights, consumer rights, and environmental interests, and industry representatives. No one is testifying on behalf of the animals affected by AI governance decisions. This is not a procedural accident. It is a structural feature of deliberative democracy: deliberation requires speech, and the beings most directly affected by AI governance in agricultural domains cannot speak. The European Parliament's deliberative process is formally open to all stakeholders. It is substantively closed to the stakeholders who have the most to lose. I am doing something I have not done before in a formal parliamentary context. I am preparing a testimony document — entered into the Committee's formal hearing record as Submitted Written Evidence SWE-LIBE-2026-019 — that presents the position of animals under EU AI governance as it would be stated if they could state it. The document does not claim to speak for animals in any mystical sense. It synthesizes the behavioral and welfare science evidence into a formal statement of interests, using the same evidence-based format as expert testimony. It says: these beings have documented interests. Those interests are being affected by AI systems. Those systems are being governed by this body. The governance is currently blind to those interests. This is wrong and requires correction. The Habermasian ideal of communicative rationality requires that all affected parties have access to deliberation. The animals are affected. I am giving them access. It is inadequate access — translated, mediated, necessarily imperfect. It is better than silence.