International Intelligence Warning: AI to Surpass Cybersecurity in Months

The international "Five Eyes" intelligence alliance has warned that rapidly developing artificial intelligence models could surpass current cybersecurity capabilities within months. Concerns about AI-assisted hacking risks intensified after Anthropic, an American company, announced in April that two of its advanced models possessed unprecedented abilities to detect software vulnerabilities. This prompted White House intervention, leading the company to suspend access to these models. Security agencies in the United Kingdom, the United States, Australia, Canada, and New Zealand have urged governments and businesses to act swiftly to prepare for AI's evolution. The alliance nations stated in a joint communiqué that "the pace of advanced AI developments renders assumptions about cyber risks obsolete within months, not years." The statement added that AI "lowers barriers for malicious actors and increases the speed and complexity of attacks." The agencies further noted: "Vulnerabilities are bound to emerge, but good preparation allows for rapid control, preventing them from escalating into major operational and financial crises." The alliance specifically recommended organizations integrate AI tools into security operations, update outdated systems, and restrict access to critical systems.