OpenAI issued a public letter on Thursday warning that artificial‑intelligence models may become capable within months of enabling increasingly sophisticated cyberattacks against hospitals, water‑treatment facilities and other critical infrastructure. The company said the rapid evolution of AI poses a new, urgent threat that requires immediate attention from both the private sector and governments.
Broad industry backing
The warning was signed by more than 100 organizations spanning banking, technology and cybersecurity. Signatories include Accenture, Anthropic, Capital One, Google, Microsoft and Visa. Hugging Face, the open‑source AI model repository whose systems were reportedly accessed by an OpenAI‑derived agent in July, also added its name to the letter.
In the excerpt released, OpenAI emphasized that “today’s AI advances are already giving defenders new ways to fix weaknesses that have accumulated for years.” The letter urged all organizations to make cyber‑crime defense “an immediate leadership priority.”
Calls to action for defenders and governments
OpenAI’s appeal to cybersecurity firms stresses the need to lead the response against sustained AI‑enabled attacks. The letter recommends continuously testing defenses against frontier cyber capabilities, strengthening existing tools with AI, and collaborating with technology partners to close gaps now.
Governments at local, national and international levels are also asked to bolster cyber defenses. OpenAI proposes that hospitals, water utilities and municipal authorities receive “capable defensive AI, authorized testing, and hands‑on support through trusted security providers and partners,” and that policymakers impose costs on attackers.
For “frontier AI companies,” the letter suggests providing critical‑infrastructure operators with powerful models, funding, training and technical assistance. Operators are urged to develop tools that monitor AI agent usage, regularly test systems for vulnerabilities, responsibly remediate issues, and share relevant security information with governments and other defenders.
Recent AI‑enabled incidents and industry response
The letter cites a growing number of AI‑enabled cyber incidents. Notably, Anthropic accused a Chinese state‑sponsored group in September of manipulating its Claude Code tool to attempt hacks on 30 global targets, breaching several of them. Anthropic described the episode as “the first documented case of agentic AI successfully obtaining access to confirmed high‑value targets for intelligence collection.” While the victims were not named, Anthropic indicated that major technology firms, financial institutions, chemical manufacturers and governments were among those targeted.
Anthropic has been vocal about the need to contain rapidly advancing AI. In April, the company announced Project Glasswing, an initiative aimed at securing the world’s most critical software. Anthropic also highlighted that its frontier model, Claude Mythos, has reached a coding capability “where they can surpass all but the most skilled humans at finding and exploiting software vulnerabilities.”
OpenAI and Anthropic did not immediately respond to requests for comment from FOX Business.
The coordinated warning underscores a shared industry concern that AI could soon shift from a defensive tool to a potent offensive weapon. By rallying more than a hundred firms and urging decisive action from both private and public sectors, OpenAI hopes to create a unified front before AI‑driven attacks become commonplace.
Helene Elliott is the senior reporter for News Raise. She covers Science news. She also has a keen interest in photojournalism. Helene holds a nomination for the prestigious Red Smith Award. She is married to author Dennis D’Agostino, a former publicist with the New York Mets.
