Prominent developers of advanced artificial intelligence tools, along with major technology firms, financial institutions, and security companies, have issued a collective warning regarding the rapid escalation of cyber threats enabled by AI. The group stressed that protecting critical infrastructure—from hospitals to technology businesses—requires immediate and decisive action, warning that the opportunity to secure systems may be fleeting.
The AI Cyber Threat and Limited Window
In an open letter released on Thursday, signatories including OpenAI, Anthropic, Google, Microsoft, CrowdStrike, Citi, and Capital One stated that their technology could soon be exploited to execute sophisticated cyberattacks against various institutions. The group emphasized that there is a “limited window” to reinforce cyber defenses, suggesting this period might last only months.
The letter noted the dual nature of AI, explaining that while the same advancements that heighten risks to public services and technology infrastructure, they also provide powerful tools for organizations to identify and “fix weaknesses” that create vulnerabilities. The signatories concluded that decisive action is needed to leverage this “defenders’ window” and make the digital world significantly more secure.
Supporting this warning, a recent cybersecurity report published by CrowdStrike indicated that AI-enabled attacks saw a substantial increase of 89% in 2025 when compared to 2024 figures.
Addressing Current Cybersecurity Deficiencies
The signatories stated that the current state of cybersecurity practices is insufficient. They cited several systemic weaknesses that leave systems exposed, including: “Longstanding bugs, excessive permissions, misconfigurations, insecure and unpatched software, weak authentication and technical debt in legacy systems.”
To counter these risks, the letter advised that security teams require significant investment and bolstering. Furthermore, defense strategies against AI cyberattacks must incorporate the most advanced AI-enabled technologies. The group also called for a greater exchange of knowledge among experts and companies, recommending that participants “Share threat intelligence and tested playbooks, and measure progress by how many organizations are protected, how quickly attacks are contained, and whether fixes work.”
Mandates for Global Action
The open letter outlined specific responsibilities for various sectors. All organizations are urged to make cybersecurity a top priority, which may necessitate replacing or upgrading older technological systems susceptible to attacks.
Specialized cybersecurity firms have a continuous duty to test defenses against evolving cyber capabilities. Governments, too, play a crucial role. They must coordinate actions and fund robust cyber defense strategies at the local, national, and international levels to contain potential threats.
Finally, the article highlighted the obligation of the leading AI companies, including those who signed the document. These frontrunners are tasked with funding training, providing “responsible” access to their models, and adequately securing their own systems.
