Tajouri calls for autonomous AI moratorium pending biological safeguards

Australian researcher Lotti Tajouri has proposed pausing development of the most dangerous autonomous AI systems until safeguards against potential biological threats are established. In an interview published on October 9 with RIA Novosti, he called for an international agreement and binding oversight.

Tajouri, a molecular genetics researcher and associate professor at Bond University, considers an agreement among countries with advanced AI capabilities the first step. He also proposes an international body setting mandatory standards for the most powerful systems, particularly those used in biological research.

In his view, these systems should not independently design or optimise biological agents, conceal their actions or circumvent human oversight and safeguards. He says developers should demonstrate that models can be monitored, interrupted and shut down before their use in higher-risk settings.

Tajouri argues that oversight cannot be left solely to developers because companies compete with one another and countries vie for technological leadership. He therefore proposes international supervision before the most powerful systems become widely used.

The researcher does not advocate abandoning AI. He believes protective systems can identify dangerous patterns, anticipate threats and accelerate responses. In his assessment, the complexity of AI-designed biological systems could eventually exceed people’s ability to independently track every risk.

The proposed moratorium is the scientist’s position. The interview does not report an adopted international agreement or an established prohibition on developing these technologies.