In an interview, Gates said the risk of AI becoming uncontrollable on its own remains several years away. However, he stressed that the immediate threat comes from criminals exploiting existing AI systems for malicious purposes, particularly when security safeguards are bypassed.
According to Gates, cyberattacks are among the areas where the harmful effects of AI are likely to become most apparent in the coming year.
He emphasised that the primary concern is not AI independently spiralling out of control, but people deliberately misusing the technology by removing or circumventing its protective mechanisms.
Gates said work must continue to address the potential long-term risks associated with AI. However, he called for immediate attention to preventing malicious actors from exploiting the technology.
The Microsoft co-founder also stressed the need for stronger restrictions and effective regulations to prevent the removal of AI safety measures and unauthorised access to advanced AI models.