More than 100 companies have signed onto an open letter calling for greater cooperation and preparation of cyber defenses across governments and other institutions. The signatories include tech giants like OpenAI, Anthropic and Microsoft, along with financial institutions, telecoms and more. The call comes as the growing capabilities of AI models lower the barrier for hackers to be able to cause disruption with critical infrastructure, important organizations and companies.
“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 says. “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 companies outline four areas where they believe action is needed now. First, the letter calls on every organization to shore up its security standards by upgrading and replacing less secure systems. Second, it says cybersecurity companies need to continuously test defenses against frontier AI capabilities and provide hands-on help to defend critical infrastructure and supply chain manufacturers. Next, it calls on governments to coordinate cyber defenses at the local, national and international levels. Finally, it says frontier AI companies should provide model access, funding, training and hands-on support for defenders.
Advertisement
Advertisement
The urgency of the warning follows months of growing concern over the increasing capabilities of the latest AI models. The first warning shot came in April, when Anthropic held its Claude Mythos model back from public release after it proved adept at finding “thousands of high-security vulnerabilities” in every major operating system and web browser. Since then, we’ve seen AI agents at OpenAI break out of their testing environment and hack another company, along with attacks on critical infrastructure — including U.S. water systems — using AI-generated exploitation scripts.
At the same time, foreign-made AI models, particularly those from China, have kept pace with models from U.S. companies. That competition has led the Trump administration to view AI as a national security priority.
