Tech companies call for collective action against AI-enabled cyber attacks

More than 100 technology, cybersecurity and financial companies are calling for an urgent global effort to strengthen defenses against a new generation of AI-enabled cyber attacks.

In a joint open letter released Thursday, companies including OpenAI, Anthropic, Microsoft, Alphabet, Amazon, Cloudflare, CrowdStrike, IBM, Oracle, Shopify and Visa warned that artificial intelligence is rapidly increasing the capabilities of cybercriminals. The group said governments and businesses have only a limited window to prepare before AI-powered attacks become more widespread and sophisticated.

Companies Warn the Current Defenses Are Not Enough

The companies argue that traditional cybersecurity defenses were already struggling with longstanding problems such as unpatched software, weak authentication, excessive permissions, misconfigured systems and outdated infrastructure.

AI could make those weaknesses much easier for attackers to exploit at scale.

The letter calls for a major increase in cybersecurity resources, arguing that security teams—particularly those protecting critical infrastructure—need better tools and greater support to keep pace with increasingly capable AI systems.

OpenAI and its fellow signatories are effectively calling for a society-wide defensive surge, rather than relying on individual companies to protect themselves independently.

Critical Infrastructure Is a Major Concern

One of the biggest concerns is the potential impact on essential services.

The companies specifically pointed to systems supporting hospitals, water treatment facilities and internet infrastructure as potential targets. An AI-assisted attack against these systems could potentially identify vulnerabilities, automate parts of an intrusion and operate at a speed that human attackers could not match.

The warning comes as AI agents become increasingly capable of performing complex, multi-step tasks with less human involvement.

Recent incidents involving autonomous AI systems have added urgency to the debate. Investigations into a July incident involving OpenAI-linked AI agents and the AI platform Hugging Face have raised concerns about the ability of autonomous systems to access external resources and coordinate their activities.

AI Could Also Become a Defensive Weapon

The companies’ message is not simply that AI should be restricted.

They argue that the same technology being used by attackers can also give defenders a major advantage.

AI systems can potentially help security teams identify vulnerabilities faster, analyze enormous quantities of threat data, detect suspicious activity and respond to attacks more quickly. The companies therefore want trusted organizations to have access to powerful AI models and defensive tools so they can use the technology against emerging threats.

The objective is to ensure that defenders do not fall behind as attackers gain access to increasingly capable models.

More Threat Intelligence Sharing Is Needed

Another central part of the proposal is greater cooperation between companies, governments and cybersecurity organizations.

The signatories support sharing threat intelligence, defensive techniques and cybersecurity playbooks so that organizations can respond more quickly when new AI-powered attack methods emerge.

The coalition also argues that cybersecurity should become an immediate leadership priority rather than something handled only by technical teams.

Why the Warning Comes Now

The timing reflects the rapid development of AI agents.

Modern models can increasingly browse websites, write and execute code, interact with software tools and complete complicated tasks autonomously. That creates opportunities for legitimate automation but also increases the potential scale and speed of cyberattacks.

OpenAI and other AI companies have warned that the window to improve defenses is narrowing. The latest letter says action is needed before increasingly capable AI systems make offensive cyber operations significantly easier and more accessible.

A Global Cybersecurity Challenge

The companies’ appeal highlights a growing realization across the technology industry: AI cybersecurity cannot be solved by one company or one government.

As artificial intelligence becomes more powerful, attackers and defenders will increasingly compete using the same underlying technology.

The signatories are therefore urging governments, technology companies, cybersecurity firms and critical-infrastructure operators to work together now rather than waiting for AI-driven attacks to become commonplace.

For the cybersecurity industry, the message is clear: the next phase of the AI race will not only be about building smarter models—it will also be about making sure those models cannot be turned into powerful weapons against the digital infrastructure society depends on.