美国Stanford大学研究人员利用生成式AI首次成功设计了完整的病毒基因组[1]。研究团队从302个AI设计方案中筛选出16种能够在实验室中有效复制的噬菌体[1]，这些病毒专门感染细菌，特别是能够有效杀死大肠杆菌[1]。
这一突破代表了生成式AI在生物设计领域的重大进展。Stanford研究人员Brian Hie表示："这是生成式AI能设计的复杂性的下一步，这是首次生成式AI被用来设计完整的基因组"[1]。所设计的噬菌体基因组约包含5400个碱基对，相比之下，最小生活细胞的基因组约有50万个碱基对[1]。
然而，这项研究也引发了严重的安全隐患。Johns Hopkins大学专家警告称这些发现提出了"紧迫的生物安全和生物安全问题"[1]，专家指出该技术存在被恶意利用设计新疾病的风险。
Researchers at Stanford University have successfully used artificial intelligence to design complete viral genomes for the first time, creating 16 novel bacteriophages capable of replicating in laboratory conditions [1]. These viruses, which infect bacteria rather than humans, were selected from 302 AI-generated designs that proved effective at killing E. coli [1].
The bacteriophage genomes designed by the AI models are approximately 5,400 base pairs in length, substantially smaller than the roughly 500,000 base pairs found in the smallest known living cell genomes [1]. According to Brian Hie, one of the researchers involved, "This represents the next step in the complexity that generative AI can design, marking the first time generative AI has been used to design complete genomes" [1].
While the technology demonstrates a significant advance in synthetic biology, experts have raised serious concerns about its implications. Specialists at Johns Hopkins University have warned that these findings present "urgent biosecurity and biosafety questions," highlighting the potential for malicious misuse in designing novel pathogens [1].