AIxBio Resources
AIxBiosecurity reading list
Selected papers and analyses on AI-enabled biological risk, catastrophic biological threats and biodefence.
00 Strategy
Biodefence strategy
The Four Pillars: A Hypothesis for Countering Catastrophic Biological Risk ↗Andrew Snyder-Beattie · Defenses in Depth · 1 October 2025
A strategic hypothesis: physical protection and early warning preserve response capacity while pathogen-specific medical countermeasures are developed.
10 big projects for reducing bio x-risk ↗Chris Bakerlee · Defenses in Depth · 9 May 2026
Ten project proposals with objectives and success criteria, including PPE stockpiles, independent equipment testing and assessment of biosecurity safeguards.
Building a Defense-in-Depth Biosecurity Strategy for the AI Era ↗RAND · 18 August 2026
Nine complementary mitigations, principally before release: information and material access controls, deterrence and proactive detection of AI-enabled biological misuse.
Industry perspective
An action plan for trusted access to biological AI, countermeasure research and early warning, with evaluation of defensive benefits and misuse risks.
Proposals linking developer oversight with synthesis screening, surveillance, respiratory protection and countermeasure capacity.
A programme of model safeguards, metagenomic surveillance and AI-assisted countermeasure development, organised around prevention, detection and response.
01 / Foundations
AI–bio risk
-
Understanding AI-Facilitated Biological Weapon Development
Centre for Long-Term Resilience · 2023 · ReportAn introduction to biological AI tools, their maturity and potential contribution to misuse, with an initial map of intervention points.
-
The near-term impact of AI on biological misuse
Centre for Long-Term Resilience · 2024 · ReportA framework for specifying and testing claims about AI-enabled uplift, distinguishing model capabilities, access conditions and threat-actor characteristics.
-
The Age of AI in the Life Sciences: Benefits and Biosecurity Considerations
National Academies of Sciences, Engineering, and Medicine · 2025 · Consensus reportAssesses biological design tools and training data, covering scientific benefits, misuse risks and options for biosecurity oversight.
Further reading · 1
-
Capability-Based Risk Assessment for AI-enabled Biological Tools
Centre for Long-Term Resilience · 2024 · FrameworkA method for assessing specialised biological tools by their capabilities and potential misuse, developed further in the Global Risk Index.
02
Risk assessment & evaluations
-
Global Risk Index for AI-enabled Biological Tools
CLTR & RAND Europe · 2025 · Risk assessmentCombines literature review and expert scoring of capability, accessibility and maturity to prioritise specialised tools for further assessment.
-
Virology Capabilities Test (VCT): A Multimodal Virology Q&A Benchmark
SecureBio-led research · 2025 · PreprintIntroduces 322 virology troubleshooting questions and specialty-matched expert baselines, measuring practical knowledge through question answering.
-
ABC-Bench: An Agentic Bio-Capabilities Benchmark for Biosecurity
SecureBio · 2026 · PreprintTests tool-using biology agents on coding-heavy tasks. Accuracy excludes refusals; separate laboratory validation involved human assistance.
Further reading · 3
-
Why we recommend risk assessments over evaluations for AI-enabled biological tools (BTs)
Centre for Long-Term Resilience · 2024 · Methods commentaryExplains why heterogeneous specialised tools require assessment beyond interactive testing, setting out the authors’ 2024 methodological position.
-
Toward Comprehensive Benchmarking of the Biological Knowledge of Frontier Large Language Models
RAND · 2025 · Research reportCompares 39 models across eight knowledge and refusal benchmarks, examining human baselines, benchmark saturation and implementation choices.
-
Do the biorisk evaluations of AI labs actually measure the risk of developing bioweapons?
Anson Ho & Arden Berg · Epoch AI · 2025 · AnalysisExamines eight labs’ disclosures and the gap between benchmark performance, physical bottlenecks and evidence of real-world misuse risk.
03
Technical safeguards
-
Strengthening nucleic acid biosecurity screening against generative protein design tools
Wittmann et al. · Science · 2025 · Research paperTests screening against AI-designed sequences and documents coordinated software updates: an empirical example of safeguard assessment leading to deployed changes.
-
A system capable of verifiably and privately screening global DNA synthesis
Baum et al. · National Science Review · 2026 · Research paperSecureDNA’s privacy-preserving screening architecture, with operational testing on commercial synthesis data and analysis of scalability and specificity.
-
Constitutional Classifiers++: Efficient Production-Grade Defenses against Universal Jailbreaks
Anthropic · 2026 · PreprintAssesses layered model safeguards through red-teaming and production-traffic evaluation, examining robustness, computational cost and false refusals.
Further reading · 2
-
Evaluating AI-assisted customer verification for synthetic nucleic acid screening
Acelas et al. · Frontiers in Bioengineering and Biotechnology · 2026 · Research paperTests AI-assisted verification in simulated screening scenarios, comparing evidence quality and workload while retaining human responsibility for order decisions.
-
Why We’re Doubling Down on Synthesis Screening
Sentinel Bio · 2026 · Prioritisation analysisCompares screening with alternative controls on physical access, focusing on provider coverage, screening quality and governance.
Active researchFourth Eon Bio ↗ — Function-based risk assessment for AI-designed biological sequences.
04
Physical biodefence
-
How I think about catastrophic biological risk — Part I
Andrew Snyder-Beattie · Defenses in Depth · 2026 · Strategic analysisExamines respiratory protection and environmental biohardening against extreme biological threats, including circumstances where pathogen-specific countermeasures may arrive too late.
-
Far-UVC (222 nm) efficiently inactivates an airborne pathogen in a room-sized chamber
Eadie et al. · Scientific Reports · 2022 · Experimental studyRoom-scale evidence for far-UVC air disinfection, measuring airborne microbial reduction rather than infection outcomes.
05
Pathogen-agnostic surveillance
-
Inferring the sensitivity of wastewater metagenomic sequencing for early detection of viruses: a statistical modelling study
The Lancet Microbe · 2025 · Modelling studyCombines empirical data from four studies to estimate detection sensitivity, linking sequence abundance with infection prevalence, incidence and sampling cost.
-
Metagenomic sequencing of composite airplane wastewater for surveillance of emerging viruses
SecureBio-led research · 2026 · PreprintCompares airplane and municipal wastewater at one airport and treatment plant, examining viral abundance and implications for surveillance sensitivity.
-
Building a three-day early-warning system for novel pathogens
SecureBio · 2026 · Programme updateA plan to shorten sample-to-result time from fourteen days to three through expanded sampling, sensitivity validation and operational changes.
06
Policy & governance
-
Defensive acceleration: the strategic pivot needed for UK biological resilience
Centre for Long-Term Resilience · 2026 · Policy reportAssesses 22 interventions and develops four programme proposals addressing surveillance, technology validation, automated screening and trusted access to frontier AI.
-
Biological Tools and the EU AI Act
Centre for Long-Term Resilience · 2025 · Policy analysisApplies general-purpose AI and systemic-risk definitions to 50 biological models, identifying ambiguities in the Act’s coverage.
-
How the UK Government should address the misuse risk from AI-enabled biological tools
Centre for Long-Term Resilience · 2024 · Policy reportProposes institutional responsibility for biological-tool risk assessment, safeguard research and responsible-development guidelines.
Further reading · 1
-
Priorities for the UK Government on frontier AI risk and Biosecurity
Centre for Long-Term Resilience · 9 September 2026 · Policy briefRecommendations include mandatory synthesis screening, defensive access to AI and restrictions on mirror-organism research.
Essential Reading
Start here for an introduction to AIxBiosecurity:
Understanding AI-Facilitated Biological Weapon Development - Centre for Long-Term Resilience
Overview of how AI capabilities intersect with biological risksThe Convergence of Artificial Intelligence and the Life Sciences - Nuclear Threat Initiative
Foundational paper on AI-bio convergence and dual-use challengesThe near-term impact of AI on biological misuse - Centre for Long-Term Resilience
Assessment of current and emerging risks from AI-enabled biologyCapability-Based Risk Assessment for AI-Enabled Biological Tools - Centre for Long-Term Resilience
Framework for evaluating risks from AI biological design toolsDeveloping Guardrails for AI Biodesign Tools - Nuclear Threat Initiative
Practical approaches to responsible development and deployment
Risk Assessments & Evaluations
Frameworks and methodologies for assessing AI-bio risks:
Global Risk Index for AI-enabled Biological Tools - Centre for Long-Term Resilience & RAND Corporation
Why we recommend risk assessments over evaluations for AI-enabled biological tools - Centre for Long-Term Resilience
Virology Capability Test (VCT) - SecureBio evaluation and paper
Assessing the Impacts of Technology Maturity and Diffusion on Malicious Biological Agent Development Capabilities - RAND Corporation
Do the biorisk evaluations of AI labs actually measure the risk of developing bioweapons? - Epoch AI
Toward Comprehensive Benchmarking of the Biological Knowledge of Frontier Large Language Models - RAND Corporation
Policy & Governance
Key policy documents and recommendations:
Biosecurity Governance Across Uncertain Artificial Intelligence Futures - RAND Corporation
How the UK Government should address the misuse risk from AI-enabled biological tools - Centre for Long-Term Resilience
Biological Tools and the EU AI Act - Centre for Long-Term Resilience
America’s AI Action Plan - U.S. Government
Emerging Technology Policy Careers - Horizon Institute for Public Service
Technical Resources
Deep dives into capabilities and methods:
Evaluating Safety and Security in Biological Research - RAND Corporation
SecureBio AI x Biosecurity Resources - Comprehensive resource hub
Issue Brief: Preliminary Taxonomy of AI-Bio Safety Evaluations - Frontier Model Forum
Secure Bio Blog - Regular updates on AI biosecurity research
Research Agendas
Strategic priorities for the field:
NTI AI-Bio Research Agenda - Nuclear Threat Initiative
Key research questions and priorities for the AI biosecurity communityBiosecurity Really: A Strategy for Victory - Hoover Institute Press
Funding Sources
Career Development and Transition Funding - Coefficient Giving
Fellowships - Future of Life Institute