Researchers in the US have used AI to design entirely new bacterial viruses (phages) by generating their complete genetic codes, successfully creating viruses that can infect and replicate within bacteria, which could help combat antibiotic-resistant infections. While this breakthrough showcases AI’s potential in biological research and medicine, it also raises safety and security concerns about the possible misuse of such technology to create harmful synthetic viruses.
Researchers in the US have, for the first time, used artificial intelligence (AI) to design entirely new viruses by creating their complete genetic codes. Unlike previous AI applications that focused on designing simple structures like antibiotics, this breakthrough involved AI models trained on genetic data from viruses, bacteria, plants, and humans to generate the genomes of bacterial phages—viruses that infect specific bacteria. Out of 300 AI-generated designs tested in the lab, 16 successfully infected and replicated within bacteria, marking a significant scientific milestone.
This development represents a major advancement in AI’s capabilities within biological research. Unlike earlier AI tools that helped design new antibiotics or vaccines, this technology can generate whole viral genomes computationally. Importantly, the viruses created target bacteria only, posing no direct risk to human health. Scientists see this as a proof of principle that AI can be used to engineer new forms of life, potentially opening doors to novel treatments and applications in medicine.
One promising application of AI-designed viruses is combating antibiotic-resistant bacterial infections. These engineered phages could serve as alternative therapies by killing bacteria that no longer respond to traditional antibiotics. However, challenges remain, such as bacteria developing resistance to these viruses as well. Experts emphasize that while this research is groundbreaking, it was pursued primarily to demonstrate AI’s potential rather than immediate clinical use.
Despite the benefits, there are significant safety and security concerns. The study highlights that it is theoretically possible to create synthetic viruses not found in nature, which could pose risks if misused. Although the current AI models cannot generate harmful human viruses, the technology could be adapted by malicious actors to develop new bioweapons. Regulatory frameworks exist to oversee such research, but they may not prevent misuse by those intent on bypassing safety protocols.
Beyond this cutting-edge research, AI’s current role in healthcare is more practical and less dramatic. AI is widely used to improve efficiency by transcribing doctors’ notes, summarizing medical information, and analyzing medical scans like mammograms to enhance screening processes. While the creation of AI-designed viruses represents a futuristic vision of biology, most AI applications in medicine today focus on supporting healthcare professionals and improving patient outcomes through data analysis and automation.