# Governable AI: Provable Safety Under Extreme Threat Models

arXiv: https://arxiv.org/abs/2508.20411 · HTML: https://mlsanity.com/paper/2508.20411/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 13 of its 56 references that are in the map.
- First posted on arXiv 28 Aug 2025.

## Builds on (most cited first, 13 of 13)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [AI Alignment: A Comprehensive Survey](https://mlsanity.com/paper/2310.19852/) (2023)
- [Alignment of Language Agents](https://mlsanity.com/paper/2103.14659/) (2021)
- [Learning Diverse and Discriminative Representations via the Principle of Maximal Coding Rate Reduction](https://mlsanity.com/paper/2006.08558/) (2020)
- [JCS: An Explainable COVID-19 Diagnosis System by Joint Classification and Segmentation](https://mlsanity.com/paper/2004.07054/) (2020)
- [A Review on Explainable Artificial Intelligence for Healthcare: Why, How, and When?](https://mlsanity.com/paper/2304.04780/) (2023)
- [Fifty Years of Prolog and Beyond](https://mlsanity.com/paper/2201.10816/) (2022)
- [AI Safety for Everyone](https://mlsanity.com/paper/2502.09288/) (2025)
- [A Dempster-Shafer approach to trustworthy AI with application to fetal brain MRI segmentation](https://mlsanity.com/paper/2204.02779/) (2022)
- [How RL Agents Behave When Their Actions Are Modified](https://mlsanity.com/paper/2102.07716/) (2021)
- [Neuro-symbolic Explainable Artificial Intelligence Twin for Zero-touch IoE in Wireless Network](https://mlsanity.com/paper/2210.06649/) (2022)
- [Actionable Guidance for High-Consequence AI Risk Management: Towards Standards Addressing AI Catastrophic Risks](https://mlsanity.com/paper/2206.08966/) (2022)
- [Automating Seccomp Filter Generation for Linux Applications](https://mlsanity.com/paper/2012.02554/) (2020)

## Abstract (arXiv)

As AI rapidly advances, the security risks posed by AI are becoming increasingly severe, especially in critical scenarios, including those posing existential risks. If AI becomes uncontrollable, manipulated, or actively evades safety mechanisms, it could trigger systemic disasters. Existing AI safety approaches-such as model enhancement, value alignment, and human intervention-suffer from fundamental, in-principle limitations when facing AI with extreme motivations and unlimited intelligence, and cannot guarantee security. To address this challenge, we propose a Governable AI (GAI) framework that shifts from traditional internal constraints to externally enforced structural compliance based on cryptographic mechanisms that are computationally infeasible to break, even for future AI, under the defined threat model and well-established cryptographic assumptions.The GAI framework is composed of a simple yet reliable, fully deterministic, powerful, flexible, and general-purpose rule enforcement module (REM); governance rules; and a governable secure super-platform (GSSP) that offers end-to-end protection against compromise or subversion by AI. The decoupling of the governance rules and the technical platform further enables a feasible and generalizable technical pathway for the safety governance of AI. REM enforces the bottom line defined by governance rules, while GSSP ensures non-bypassability, tamper-resistance, and unforgeability to eliminate all identified attack vectors. This paper also presents a rigorous formal proof of the security properties of this mechanism and demonstrates its effectiveness through a prototype implementation evaluated in representative high-stakes scenarios.