Genomics researcher Lotti Tajouri says one or two irresponsible users could pose a serious threat if no safeguards stop misuse. He urged risk assessments to account for wider access to AI-enabled research tools.
Even one or two people with access to powerful artificial-intelligence systems could create a serious biological threat if an attempt at misuse went unchecked, Australian genomics and molecular-biology specialist Lotti Tajouri told RIA Novosti.
Risk assessments, Tajouri said, should not assume every user will act responsibly. He also warned that AI could expand the number of people and laboratories able to analyse large volumes of biological data and accelerate research design.
As context, RIA Novosti cited Stanford researchers’ work with the Evo 1 and Evo 2 models. In laboratory tests, researchers synthesised and tested nearly 300 model-designed bacteriophages. Sixteen were able to attack E. coli. Bacteriophages are viruses that infect bacteria, so the result does not mean that a virus dangerous to humans was created.
Tajouri said that as synthetic biology becomes faster and more accessible, risk assessment and oversight should account for deliberate misuse. His warning concerns the potential expansion of access, not evidence that a new biological threat has already emerged.