Building Barricades: Red-Teaming Agentic Biosynthesis Planning
Abstract
The rapid democratization of biotechnology, enhanced by large language models (LLMs) and agentic software, is eroding longstanding barriers that have limited biosynthesis planning to experts. We present BioBulbul, an agentic system that couples LLMs with open-source biodesign tools for retrobiosynthesis, enzyme selection, gene cluster design, and DNA assembly planning. Using simple natural language prompts, BioBulbul autonomously generates plausible biosynthesis pathways and corresponding gene cluster sequences for several known small molecule toxins, demonstrating that non-experts could feasibly access them using an agentic system. Our red-teaming experiments confirm the technical viability of LLM-driven toxin biosynthesis while highlighting three principal barriers: scarcity of open-source biodesign tools, limited structured enzymatic reaction data, and the complexity of many small molecule toxin pathways. We argue that these barriers are likely to diminish within a decade, expanding the biological threat landscape. To mitigate emerging risks, we recommend expanding DNA synthesis screening to include chokepoint enzymes, extending chemical supplier controls to precursor compounds, establishing continuous AI-assisted threat surveillance pipelines, and developing robust LLM guardrails. Coordinated investment across government, industry, and AI developers is essential to safeguard against the misuse of agentic biosynthesis planning.
Related
- Floral Volatile Organic Compounds of Chinese Cymbidium Orchids: From Biosynthesis and Function to Genetic Regulation
- Genomics of the Genus Rhodiola: From Genome Assembly to High-Altitude Adaptation and Bioactive Compound Biosynthesis
- Structure-Based Drug Design of Novel Inhibitors Targeting the Thiamine Biosynthesis Pathway in Bacteria
- Mycobacterium tuberculosis requires riboflavin biosynthesis for survival in the host
- Hypoxia and ascorbic acid: two interconnected microenvironmental factors driving collagen biosynthesis
BibTeX
@misc{roehner2026building,
title = {{Building Barricades: Red-Teaming Agentic Biosynthesis Planning}},
author = {Nicholas Roehner and Bryan Bartley and Jacob Beal and Caitlin Jagla and Troy Timmerman and Kemper Talley},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.08218},
eprint = {2611.08218},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.08218},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Building Barricades: Red-Teaming Agentic Biosynthesis Planning
AU - Nicholas Roehner
AU - Bryan Bartley
AU - Jacob Beal
AU - Caitlin Jagla
AU - Troy Timmerman
AU - Kemper Talley
PY - 2026
DA - 2026/11/05
PB - arXiv
DO - 10.48550/arxiv.2611.08218
UR - https://arxiv.org/abs/2611.08218
AB - The rapid democratization of biotechnology, enhanced by large language models (LLMs) and agentic software, is eroding longstanding barriers that have limited biosynthesis planning to experts. We present BioBulbul, an agentic system that couples LLMs with open-source biodesign tools for retrobiosynthesis, enzyme selection, gene cluster design, and DNA assembly planning. Using simple natural language prompts, BioBulbul autonomously generates plausible biosynthesis pathways and corresponding gene cluster sequences for several known small molecule toxins, demonstrating that non-experts could feasibly access them using an agentic system. Our red-teaming experiments confirm the technical viability of LLM-driven toxin biosynthesis while highlighting three principal barriers: scarcity of open-source biodesign tools, limited structured enzymatic reaction data, and the complexity of many small molecule toxin pathways. We argue that these barriers are likely to diminish within a decade, expanding the biological threat landscape. To mitigate emerging risks, we recommend expanding DNA synthesis screening to include chokepoint enzymes, extending chemical supplier controls to precursor compounds, establishing continuous AI-assisted threat surveillance pipelines, and developing robust LLM guardrails. Coordinated investment across government, industry, and AI developers is essential to safeguard against the misuse of agentic biosynthesis planning.
ER -