Artificial intelligence is unlikely to significantly alter the threat posed by bioweapons or viruses, according to technology reporters. Historical incidents, such as Aum Shinrikyo's failed attempts to use botulism and anthrax in Tokyo, highlight the logistical difficulties involved in such attacks.
While humans could make poor decisions regarding AI's role in nuclear systems, this is viewed more as human error than a rogue AI issue. Some experts caution against anthropomorphising AI or attributing intention to it, suggesting this plays into industry narratives pushing for favourable regulation.
Despite claims of benefits from chatbots to scientific applications like Google's AlphaFold model for protein prediction, the wider global economic advantages predicted by the AI industry have not yet fully emerged. These benefits often rely on AI agents becoming more active in the workforce, a stage not yet reached.