textdi.com - A powerful AI model called Claude Mythos, developed by Anthropic, is too dangerous to be released to the public due to its exceptional cyber hacking capabilities. Although a limited number of large US companies and 40 other critical infrastructure providers have access to the model for defensive purposes, experts warn that Australia is vulnerable to cyber attacks due to the country's dependence on foreign software and infrastructure. The issue is that AI's power to find and exploit vulnerabilities is increasing at a rapid rate, and Australia needs to be proactive in regulating and mitigating these risks to prevent potential cyber disasters.

New AI threat looms but Australian firms don't have access needed to prepare

A powerful AI model called Claude Mythos, developed by Anthropic, is too dangerous to be released to the public due to its exceptional cyber hacking capabilities. Although a limited number of large US companies and 40 other critical infrastructure providers have access to the model for defensive purposes, experts warn that Australia is vulnerable to cyber attacks due to the country's dependence on foreign software and infrastructure. The issue is that AI's power to find and exploit vulnerabilities is increasing at a rapid rate, and Australia needs to be proactive in regulating and mitigating these risks to prevent potential cyber disasters.