Human-AI Safety: A Descendant of Generative AI and Control Systems Safety

TL;DR

This paper explores the intersection of AI safety and control systems, emphasizing the importance of understanding human-AI feedback loops for ensuring safety in advanced AI systems.

Contribution

It introduces a unifying formalism for modeling dynamic human-AI interactions and outlines a technical roadmap for next-generation human-centered AI safety.

Findings

Highlighting the entanglement of AI outputs with human responses over time

Identifying open challenges in current AI safety approaches

Proposing a formal framework for analyzing human-AI safety interactions

Abstract

Artificial intelligence (AI) is interacting with people at an unprecedented scale, offering new avenues for immense positive impact, but also raising widespread concerns around the potential for individual and societal harm. Today, the predominant paradigm for human--AI safety focuses on fine-tuning the generative model's outputs to better agree with human-provided examples or feedback. In reality, however, the consequences of an AI model's outputs cannot be determined in isolation: they are tightly entangled with the responses and behavior of human users over time. In this paper, we distill key complementary lessons from AI safety and control systems safety, highlighting open challenges as well as key synergies between both fields. We then argue that meaningful safety assurances for advanced AI technologies require reasoning about how the feedback loop formed by AI outputs and human behavior may drive the interaction towards different outcomes. To this end, we introduce a unifying formalism to capture dynamic, safety-critical human--AI interactions and propose a concrete technical roadmap towards next-generation human-centered AI safety.