The Rise of AI in Biological Research: Opportunities and Risks

In February 2026, OpenAI and Ginkgo Bioworks announced that OpenAI's GPT-5 had autonomously designed and executed 36,000 biological experiments using a robotic cloud laboratory. This facility allows automated equipment to conduct experiments remotely, significantly reducing the cost of producing proteins by 40%. This advancement marks a shift in biological research from traditional methods to a more engineering-like approach, where AI can design, build, test, and iterate on biological systems rapidly. AI's ability to explore thousands of design variations in parallel accelerates protein engineering, potentially leading to faster responses to emerging infections and more affordable drugs.

The Conversation

Despite the benefits, the integration of AI in biological research raises significant concerns, particularly regarding the dual-use problem. Technologies developed for beneficial purposes can also be repurposed for harmful applications. Researchers have identified that AI models can optimize viral spread without specialized training, posing risks for bioweapon development. Current oversight mechanisms are deemed inadequate to address these emerging threats. A risk-scoring tool has been developed to evaluate how AI could modify a virus's capabilities, highlighting the potential for misuse.

The Conversation

Recent studies have produced conflicting findings regarding AI's role in enabling individuals with limited biology training to conduct complex lab work. A study by Scale AI and SecureBio found that novices using AI tools could complete biosecurity-related tasks with four times greater accuracy than trained experts. Conversely, research from Active Site indicated that while AI assistance improved task completion rates, it did not significantly enhance novices' ability to produce viruses in a biosafety laboratory. This discrepancy underscores the ongoing debate about the implications of AI in biological experimentation.

The Conversation

As AI systems increasingly run experiments autonomously, existing regulations lag behind. Current biological research rules do not account for AI-driven automation, and AI regulations do not specifically address its application in biology. The Biden administration's 2023 executive order on AI security included biosecurity provisions, but these were revoked under the Trump administration. A bipartisan bill introduced in 2026 aims to mandate DNA screening, yet it does not address AI-designed sequences that could evade detection. The lack of provisions in the 1975 Biological Weapons Convention further complicates governance.

The Conversation

Experts emphasize the need for coordinated government action to address the risks associated with AI in biological research. The U.K. AI Security Institute and the U.S. National Security Commission on Emerging Biotechnology have called for improved safety evaluations and governance of biological data. While some AI companies, like Anthropic and OpenAI, have begun implementing voluntary safety measures, the pace of AI development may outstrip the ability of any single entity to assess risks effectively.

The Conversation

AI holds the potential to revolutionize biological research, enabling rapid advancements in drug development and disease response. However, the risks associated with its misuse necessitate careful consideration and proactive governance. Striking a balance between fostering innovation and ensuring safety is critical as the field navigates the complexities of AI-driven biology.

The Conversation