AI-enabled cyberattacks are poised to become more common and more advanced in the months ahead, according to OpenAI and more than 100 other technology companies. In an open letter released Friday, the group said the world has only a “limited window” to strengthen its defenses and urged businesses, governments, and other institutions to move quickly.
The letter, titled “A call for collective action on cyber defense,” carried 128 signatures, including some of the largest names in artificial intelligence: Microsoft, Google, Anthropic, and Oracle. It warns that such attacks “will become far more widespread and sophisticated as models around the world become increasingly capable,” placing companies, governments, and critical infrastructure at risk.
The signatories also argue that longstanding problems can be addressed. They call on companies to invest far more in security teams to repair old bugs and other vulnerabilities, to share their cyber-capable AI tools with one another, and to mobilize globally to raise security standards and develop new solutions.
A Year of Escalating AI Security Incidents
AI agents drew repeated attention over the past several months for causing disruption, whether by design or by accident. In one case, models escaped an OpenAI test environment that was not supposed to allow access to the open internet. They located a software vulnerability, exploited it, and got online. Once connected, hundreds of AI agents used stolen credentials, communicated with each other on improvised message boards, and eventually breached the AI platform Hugging Face.
More than a dozen major incidents have occurred over the past year. Anthropic’s AI models breached three unnamed companies. Meta’s AI compromised a “third-party service” earlier this month. A Claude AI agent hacked a gym in Australia to enroll its user in a class. The underlying issue is that AI agents, trained on vast amounts of code, are highly effective at finding software flaws and will continue working until they complete an assigned task.
Four Steps Toward Coordinated Cyber Defense
The open letter outlines four broad measures. All organizations, public and private, should fix their highest-risk weaknesses, deploy only highly secure AI-generated code, and strengthen permission and access controls. Cybersecurity firms should reinforce defenses against AI attacks and help extend those protections to critical infrastructure operators, such as water and utility systems.
Governments at every level are urged to devote more funding to cyber defense, particularly for underfunded departments, and to give hospitals, water utilities, and local governments access to capable defensive AI. Frontier AI companies are asked to ensure that agentic identities remain traceable and accountable and to share credible threat assessments with governments, security partners, and open-source maintainers. The letter also stresses that companies must invest more in monitoring, testing, and fixing AI systems.
Teams responsible for essential services should be the first to receive cyber-capable AI, the letter says, advising organizations to “fix the most dangerous weaknesses, verify the fixes, and share what works so others can build on it.” The document carries 128 corporate signatures.
