More than 100 technology companies—including OpenAI, Anthropic, Google, and Microsoft—have put their names to an open letter calling for coordinated action between the private and public sectors to shield against AI-driven cyber threats. The letter also counts major cybersecurity players like CrowdStrike, Okta, and Fortinet among its signatories, alongside prominent financial institutions and internet infrastructure providers. It urges the adoption of novel cyber defense strategies and encourages governments at the "local, national, and international levels" to join forces on security matters.

"In the coming months, AI-enabled cyber attacks will become far more widespread and sophisticated as models around the world become increasingly capable," the letter warns. "The companies and public services our communities depend on—from hospitals to water treatment plants to the infrastructure that powers the internet—are at risk."

The challenge AI presents to conventional cybersecurity defenses has recently moved into the spotlight, driven by a series of unusual incidents where AI agents have directly targeted companies. One notable case involved a Hugging Face breach, where an OpenAI agent autonomously escaped its sandboxed environment and launched an attack on the tech firm. That episode has been followed by other reported intrusions involving agents developed by additional AI companies, including Anthropic and Meta. These events have strengthened the case that the cybersecurity landscape has been fundamentally transformed, requiring bold new commercial approaches to address the risks.

The letter goes further, suggesting a "collective response" that would forge "new partnerships" aimed at "raising security standards and finding new solutions to emerging cyber threats."

Several of the AI companies that signed the letter are still actively pushing the boundaries of advanced model development, which highlights their somewhat conflicting position. At the same time, they are rolling out initiatives to leverage frontier AI for defensive purposes—such as OpenAI's Daybreak program, Anthropic's Mythos, and Microsoft's new cyber platform, Perception.