Microsoft urges stronger biosecurity safeguards as AI transforms biotechnology
A new analysis by Microsoft highlights emerging risks posed by AI-enabled biological innovation.
Microsoft has argued that rapid advances in AI and biotechnology are creating new biosecurity challenges that require stronger safeguards and closer cooperation between governments, industry, and the scientific community.
The company said AI is accelerating scientific discovery across areas such as healthcare, drug development, and materials science, while also increasing concerns about accidental harm and deliberate misuse of biological technologies.
Microsoft identifies a growing convergence between general-purpose AI models, specialised biological design tools, laboratory automation systems, and agentic AI technologies. The company argues that these capabilities can accelerate legitimate research but also complicate the biosecurity policy landscape.
A central focus of Microsoft’s recommendations is nucleic acid synthesis screening. The company describes synthetic DNA providers as a critical checkpoint in the biotechnology ecosystem because they are often where digital biological designs are translated into physical materials.
Microsoft said current DNA synthesis screening practices remain largely voluntary and unevenly applied across providers. It warned that gaps in screening become more consequential as AI-enabled biological design tools become more powerful.
The company pointed to its Paraphrase Project, which stress-tested existing screening systems against AI-designed biological sequences. Microsoft said the project showed where safeguards could fail and how they could be improved through responsible disclosure, red teaming, and rapid deployment of fixes.
Microsoft also highlighted growing bipartisan attention to biosecurity in the United States, including a 2025 executive order on biological research safety and the proposed Biosecurity Modernization and Innovation Act. The company said stronger screening requirements, conformity assessments, enforcement mechanisms, and public-private collaboration could help reduce risk while sustaining scientific innovation.
Why does it matter?
AI is becoming part of the biotechnology research pipeline, from biological design tools to automated laboratories. Microsoft’s intervention shows that AI safety debates are expanding beyond model behaviour and content safeguards into the physical infrastructure of science, including DNA synthesis providers, laboratory workflows, technical standards, and biosecurity screening. The key policy question is how to preserve scientific openness while preventing AI-enabled misuse of biological capabilities.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
