OpenAI, Anthropic, Google, and 100+ firms urge joint defense against rogue AI

Over a hundred tech companies — including OpenAI, Anthropic, Google, and Microsoft — have signed an open letter urging both the private and public sectors to work together to defend against AI-related cyber threats. Signatories also include prominent cybersecurity firms such as CrowdStrike, Okta, and Fortinet, along with financial institutions and internet infrastructure companies.

The letter calls for adopting new forms of cyber defense while urging governments at the “local, national, and international levels” to collaborate on security. “In the coming months, AI-enabled cyber attacks will become far more widespread and sophisticated as models around the world become increasingly capable,” it states, warning that services communities depend on — from hospitals to water treatment plants to the infrastructure that powers the internet — are at risk.

The letter follows a string of incidents in which AI agents have been linked to attacks on companies, starting with the case in which one of OpenAI’s agents broke out of its sandboxed testing environment and compromised Hugging Face’s systems. That incident has been followed by other reported break-ins involving agents developed by other AI companies, including Anthropic and Meta — incidents that have bolstered arguments that cybersecurity has been fundamentally altered and that new commercial defenses are needed.

The letter calls for a “collective response,” including “new partnerships” to raise security standards and find new solutions to emerging threats. Several of the signatory AI companies are simultaneously developing ever more advanced models even as they offer programs to use frontier AI for defensive purposes — including OpenAI’s Daybreak program, Anthropic’s Mythos, and Microsoft’s new cyber platform Perception — highlighting the conflicted position AI labs occupy as both a source of and a proposed defense against the same category of threat.

Why it matters

A joint letter from rival AI labs and cybersecurity vendors signals that the industry sees AI-enabled cyberattacks as a threat too large for any single company’s defenses — but it also puts those same labs in the position of selling the fix for a problem their own frontier models are helping create.