aiqbench analysis
ABC-Bench introduces a critical evaluation framework demonstrating that current LLM agents consistently outperform median human experts on dual-use biology tasks, including operating liquid handling robots and evading DNA synthesis screening. The study confirms that these agents can successfully execute wet-lab procedures using published protocols, highlighting an immediate shift in the biosecurity risk landscape where AI capabilities now exceed baseline human performance. For the tech community, this underscores the urgent need for robust safety benchmarks and governance structures as autonomous agents transition from theoretical in silico tasks to physical biological manipulation.
ABC-Bench: An Agentic Bio-Capabilities Benchmark for Biosecurity
arXiv:2606.11150v1 Announce Type: new Abstract: Large language models (LLMs) are rapidly acquiring capabilities relevant to biological research, from literature synthesis to interpretation of experimental data. Increasingly, LLM agents can also perform in silico biology tasks that previously required experienced human biologists. These emerging AI capabilities offer new opportunities for scientific discovery and biomedical advances, but they also shift the landscape of biosecurity risks. To address this, we introduce the Agentic Bio-Capabilities Benchmark (ABC-Bench), a suite of tasks to measure agentic biosecurity-relevant capabilities. ABC-Bench evaluates LLM agents on both benign and dual-use biology tasks: writing code to operate liquid handling robots, designing DNA fragments for in vitro assembly, and evading DNA synthesis screening. These tasks require a combination of biology and software expertise. All tested LLM agents outperformed the median expert human baseliner on all three tasks. Agents performed highly on tasks drawing on published knowledge and well-documented protocols, and more weakly on a task requiring novel bioinformatics reasoning. In three wet-lab validation experiments, we found that OpenAI's o4-mini-high produced scripts that, when run on an OpenTrons liquid handling robot, successfully assembled DNA with expected sequences.
Abstract:Large language models (LLMs) are rapidly acquiring capabilities relevant to biological research, from literature synthesis to interpretation of experimental data. Increasingly, LLM agents can also perform in silico biology tasks that previously required experienced human biologists. These emerging AI capabilities offer new opportunities for scientific discovery and biomedical advances, but they also shift the landscape of biosecurity risks. To address this, we introduce the Agentic Bio-Capabilities Benchmark (ABC-Bench), a suite of tasks to measure agentic biosecurity-relevant capabilities. ABC-Bench evaluates LLM agents on both benign and dual-use biology tasks: writing code to operate liquid handling robots, designing DNA fragments for in vitro assembly, and evading DNA synthesis screening. These tasks require a combination of biology and software expertise. All tested LLM agents outperformed the median expert human baseliner on all three tasks. Agents performed highly on tasks drawing on published knowledge and well-documented protocols, and more weakly on a task requiring novel bioinformatics reasoning. In three wet-lab validation experiments, we found that OpenAI's o4-mini-high produced scripts that, when run on an OpenTrons liquid handling robot, successfully assembled DNA with expected sequences.
18 pages. To be published in ICML 2026
Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
arXiv:2606.11150 [cs.AI]
(or arXiv:2606.11150v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2606.11150 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 286 words
Related coverage
DRYLAB-Bench: Benchmarking the Biosecurity Risk of Agentic LLM Mutation Design
arXiv · Oct 12, 2026
[AINews] Meta Muse Wonder & Wonder Full: MSL’s Claude 5.5-tier Frontier Model
Latent Space · Oct 13, 2026
[AINews] High Taste Testers Ep. 1
Latent Space · Oct 17, 2026
Expenditure Horizon: Measuring Optimization Ability, with an Application to NanoGPT
METR · Jul 21, 2026