The Reality of AI and Biorisk

2024-12-02

AI Summary

The AI-biorisk field suffers from a lack of rigorous theoretical frameworks, weak empirical foundations, methodological flaws, and insufficient transparency—hindering reliable assessment of biological threats posed by large language models (LLMs) and AI-enabled bio-tools. Method: This study conducts the first interdisciplinary critical review and threat modeling analysis—integrating AI safety, synthetic biology, and risk assessment—to systematically evaluate two critical threat pathways: information acquisition/planning and de novo synthesis of novel biological entities. Contribution/Results: We find that current LLMs and AI bio-tools do not constitute an immediate biosecurity risk; however, significant methodological limitations severely undermine the reliability of risk predictions. To address this, we propose a research roadmap emphasizing empirical rigor, reproducibility, and threat-targeted evaluation. This work establishes foundational theoretical principles and methodological scaffolding for developing scientifically grounded, operationally viable AI-biorisk assessment frameworks.

Technology Category

• Philosophy and Ethics of AI: Safety, Robustness & Trustworthiness
• Natural Language Processing: Safety and Robustness
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Security and Privacy: Security and privacy of machine learning and AI applications

Abstract

To accurately and confidently answer the question 'could an AI model or system increase biorisk', it is necessary to have both a sound theoretical threat model for how AI models or systems could increase biorisk and a robust method for testing that threat model. This paper provides an analysis of existing available research surrounding two AI and biorisk threat models: 1) access to information and planning via large language models (LLMs), and 2) the use of AI-enabled biological tools (BTs) in synthesizing novel biological artifacts. We find that existing studies around AI-related biorisk are nascent, often speculative in nature, or limited in terms of their methodological maturity and transparency. The available literature suggests that current LLMs and BTs do not pose an immediate risk, and more work is needed to develop rigorous approaches to understanding how future models could increase biorisks. We end with recommendations about how empirical work can be expanded to more precisely target biorisk and ensure rigor and validity of findings.

Problem

• Biological Risk
• Artificial Intelligence
• Language Models

Innovation

• AI biosecurity risks
• language models
• risk mitigation framework