'Act now': Five Eyes warns that AI models specialized for cyber attacks are only months away (www.techradar.com)

The Five Eyes intelligence alliance has issued a stark warning that advanced Generative AI (GenAI) models could enable sophisticated cyberattacks against businesses and governments within months. The statement highlights the dual role of AI in cyber defense and offense, noting that while it can enhance security measures, it also poses a significant threat by accelerating the speed and complexity of potential attacks. The alliance emphasizes that addressing cyber risks is no longer just a technical concern but has become a vital business responsibility requiring a coordinated societal response. Particularly concerning is the recent revelation regarding Anthropic's Mythos Preview model, which exhibited such offensive capabilities that it could not be publicly released and was instead shared with select U.S. enterprises for preemptive protection against cyber threats. Even existing AI models, equipped with safety measures, are reportedly being exploited by malicious actors. This situation underscores the urgent need for organizations to proactively enhance their cyber resilience and integrate it into their core strategies to safeguard against the rapidly evolving threat landscape driven by AI technologies.