The article introduces the "HardFlow" algorithm, a new method designed to improve the reliability of generative AI models in safety‑critical contexts. By enforcing strict requirements, HardFlow aims to ensure that the AI’s outputs meet high‑quality standards even when approximate solutions are insufficient.
arXiv:2502. 12445v2 Announce Type: replace Abstract: AI safety is a rapidly growing area of research that seeks to prevent the harm and misuse of frontier AI technology, particularly with respect to generative AI (GenAI) tools that are capable of creating realistic and high-quality content through text prompts.
arXiv:2606. 05952v1 Announce Type: cross Abstract: In this work, we propose an agentic gamification framework for hazard-informed learning of robot safety policies through synthetic scenarios.
In this work, we propose an agentic gamification framework for hazard-informed learning of robot safety policies through synthetic scenarios. We model scenario generation as an adversarial game between two agents: a Red Team that explores the space of potential failures by constructing hazardous situations, and a Blue Team that incrementally refines safety policies to prevent them.
arXiv:2605.29713v2 Announce Type: replace-cross Abstract: This book provides a compact, derivation-oriented introduction to the mathematical foundations of modern generative artificial intelligence....
arXiv:2602. 16065v2 Announce Type: replace-cross Abstract: As artificial intelligence (AI)-generated content proliferates, models are increasingly trained on their own outputs, risking progressive degradation or collapse.
Ensuring that AI systems are built, deployed, and used safely is critical to our mission.
We (along with researchers from Berkeley and Stanford) are co-authors on today’s paper led by Google Brain researchers, Concrete Problems in AI Safety. The paper explores many research problems around ensuring that modern machine learning systems operate as intended.
The article "Math for AI safety: an invitation for mathematicians" calls for new mathematical tools to ensure AI remains understandable, controllable, and cooperative. It outlines specific mathematical fields—logic, game theory, probability, algebra, representation theory, analysis, and geometry—each paired with an open problem tailored for mathematicians without AI safety background. The piece invites researchers to contribute to designing AI that is legible, steerable, and aligned with human values.
arXiv:2608. 12325v1 Announce Type: new Abstract: Autonomous reasoning is among the most scientifically and economically motivating topics in AI today.
Discover how OpenAI's new safe-completions approach in GPT-5 improves both safety and helpfulness in AI responses—moving beyond hard refusals to nuanced, output-centric safety training for handling dual-use prompts.
arXiv:2608. 20053v1 Announce Type: new Abstract: The integration of Artificial Intelligence (AI) in safety-critical aviation systems presents significant challenges for certification and deployment.
The article surveys recent progress in tractable probabilistic generative modeling, with a focus on Probabilistic Circuits (PCs). It offers a unified view of the trade‑offs between expressivity and tractability, outlining design principles, algorithmic extensions, and a taxonomy of the field. The review also covers deep and hybrid PCs that integrate ideas from deep neural models, and highlights challenges and open questions for future research.