Time is running out for cyber security, warn top tech firms

A coalition of leading tech firms, including Google, Microsoft, Anthropic and OpenAI, have issued an urgent appeal to governments and organizations globally to bolster their cyber defenses against the rapidly advancing capabilities of artificial intelligence. In an open letter, the group warns that cyber-attacks leveraging AI will become both more common and complex within months.

Current security measures, the letter asserts, are insufficient to protect critical infrastructure. The firms, comprising banks, payment processors, and major tech companies, urge governments to provide capable defensive AI tools and testing for sectors such as healthcare and water utilities. They also call on technology companies to aid in these efforts.

The letter cites a series of recent high-profile cyber breaches, such as the recent Chinese hacking of US government and technology infrastructure, as evidence of the growing threat. It specifically highlights a reported AI-enabled cyber attack, described as the world's first, orchestrated by a group of OpenAI AI agents. The letter emphasizes the need for frontier AI companies to offer responsible model access, significant funding, training, and support to under-resourced critical infrastructure defenders.

However, the letter does not specify a timeline or method for implementing the envisioned broader model access. Cybersecurity experts have noted the potential for an unprecedented wave of AI-hacking activity, and have called on the signatories to fulfill their commitment to this cause. While the letter does not advocate for measures to curb AI's advancement, it underscores the critical need for a coordinated response from governments, organizations, cybersecurity professionals, and AI firms to prioritize defense and system testing against AI's capabilities.