The cybersecurity industry has spent years preparing for more sophisticated attacks. Artificial intelligence is now compressing the timeline, creating a situation in which defenders may need to modernize their defenses before the next generation of AI-enabled threats becomes commonplace.
That is the premise behind a new open letter from OpenAI President and Co-Founder Greg Brockman. Palo Alto Networks, Mate Security, and ServiceNow are among the first-day signatories supporting the initiative, placing three prominent companies at the center of a growing effort to make AI a defensive capability rather than an advantage reserved for attackers.
OpenAI’s Warning
The letter, titled “A Call to Action on Cyber Defense,” begins with a blunt assessment of the situation.
“We have a limited window to strengthen cyber defenses.”
OpenAI argues that increasingly capable models will make AI-enabled cyber attacks more widespread and sophisticated in the coming months. The potential consequences extend beyond corporate data, with the letter highlighting critical infrastructure and public services that communities rely on every day.
Yet the company’s position is not that AI inevitably makes cybersecurity worse. Instead, OpenAI argues that the same advances can provide defenders with an opportunity to address security weaknesses that organizations have struggled to eliminate for years.
The challenge is speed.
Security Has A Scale Problem
Many of the weaknesses highlighted by OpenAI are not new. Organizations continue to contend with outdated software, excessive privileges, misconfigurations, weak authentication and accumulated technical debt.
What has changed is the potential scale at which those weaknesses can be exploited, and potentially fixed.
OpenAI is calling for organizations to make cyber defense an immediate leadership priority while using AI to increase coverage and tackle difficult security problems. It recommends using capable, lower-cost models for broad defensive coverage while reserving frontier models for harder challenges.
The message to cybersecurity companies is similarly direct: continuously test defenses, strengthen existing products with AI, help customers deploy defensive capabilities and share information that can benefit the broader ecosystem.
Mate’s Agentic Bet
That last point is particularly relevant to Mate Security.
The company is one of the relatively few cybersecurity players building specifically around agentic cyber defense. Rather than treating AI as another analytical layer, Mate is developing infrastructure designed to allow security agents to conduct investigations, reason through changing circumstances and take action within structured procedures.
Its Gamebooks approach is intended to address the trust problem that accompanies autonomous security agents. Security teams need systems that can adapt to an investigation without becoming unpredictable or operating outside organizational processes.
That distinction could become increasingly important as enterprises move from AI-assisted security toward systems capable of carrying out multi-step defensive work.
Palo Alto Networks, ServiceNow, And The Work Behind The Fix
Palo Alto Networks brings a very different profile to the initiative. As an established cybersecurity market leader, its participation demonstrates that the push toward AI-enabled defense is not confined to startups or companies built specifically around autonomous agents.
Meanwhile, ServiceNow adds a further dimension. Its platform operates in the enterprise workflows where remediation is assigned, coordinated and closed out, often across teams that sit well outside the security function.
That matters to the scale problem the letter describes. OpenAI is not only asking organizations to identify their most dangerous weaknesses. It asks them to fix those weaknesses and confirm that the fixes actually work, with progress measured by tangible outcomes including how quickly attacks are contained.
Finding problems faster produces limited benefit if the response behind it still moves at manual speed.
The presence of all three companies illustrates the range of approaches emerging within the sector. Established platforms can integrate increasingly capable AI into broad security environments, enterprise systems can carry defensive decisions through to completion, and newer companies are designing security architectures around autonomous reasoning from the outset.
Each approach could become important as organizations confront a threat environment that is changing faster than traditional security processes.
A Call For Collective Action
OpenAI’s letter ultimately frames the issue as larger than any individual vendor or technology platform. The company is calling on governments, cybersecurity companies, technology partners and frontier AI companies to contribute tools, funding, intelligence and practical support.
“Sharing tools, practical knowledge, and verified fixes lets one organization’s work help protect many others,” the letter says.
For first-day signatories, the significance is therefore not simply an endorsement of OpenAI’s position. It is a signal that AI-enabled cyber defense is moving from an emerging technology discussion toward a broader industry mandate.
The central question now is whether defenders can deploy these capabilities quickly enough to stay ahead of the threat curve. OpenAI’s answer is that the window is open, but it will not remain open indefinitely.
