AI Constructs World's First Replicating Virus

Researchers in the United States have utilized Artificial Intelligence (AI) to engineer an entirely novel class of viruses capable of replicating inside a laboratory setting. This marks the historic first instance where a complete viral genome was designed entirely through generative AI frameworks. The 16 newly generated viruses target and infect bacteria specifically, posing no health threats to humans. While scientists have hailed the milestone as a revolutionary leap that could transform disease treatment, experts caution that AI's newly demonstrated capability to draft viable biological organisms raises urgent biosecurity concerns.

Commenting on the achievement, Assistant Professor Brian Hie of Stanford University stated, "This represents the next frontier in designing complex biological structures using generative AI. It is the first time a fully functional genome capable of replication and intracellular activity has been generated from scratch by AI. This is uncharted territory for us." The underlying technology operates similarly to Large Language Models (LLMs) like ChatGPT; however, instead of predicting natural language text, the system predicts the fundamental code of life: genetic sequences. AI models named Evo-1 and Evo-2 were trained on extensive genetic data spanning viruses, bacteria, plants, and humans to design specialized bacterial-targeting viruses known as bacteriophages. Out of 302 candidate designs generated by the AI, researchers synthesized 16 functional strains that successfully destroyed E. coli bacteria. Samuel King, a PhD researcher at Stanford, shared that clear rings appeared in bacteria-filled Petri dishes during late-night tests, confirming that the synthetic phages were actively eliminating the bacterial colonies.

The ability to generate customized phages presents a promising alternative against antibiotic-resistant superbug infections, marking a massive leap forward for synthetic biology. Nevertheless, the scientific community expresses grave concerns over potential dual-use misapplications, where the same methodology could be exploited to design human pathogens. Writing in the journal 'Science', Dr. Thomas Inglesby and Dr. Moritz Hancke from the Johns Hopkins Center for Health Security noted that the research prompts "urgent questions regarding biosecurity and biosafety." Experts emphasize that the debate is no longer about whether AI can design functional viruses, but rather how global frameworks can strictly regulate the technology to mitigate catastrophic biological risks, specifically prohibiting research into human-infecting pathogens.

//DBTech/BBC/SME//